413 replies to this topic

[oldradagast]

ScarecrowES, on 06 May 2016 - 10:44 AM, said:

The only way you get out of the Blackjack getting bigger is if you eyeball every mech... you don't apply a standard and look at mechs subjectively. So basically balance by Paul.

Agreed. Math is the safest place to start as a baseline for proper mech scaling. Anything else, as you said, results in "balance by Paul / eyeballing / guessing / gut feeling." Not only has PGI demonstrated a poor track record on that, I'll be honest and say I have serious doubts much of anyone - myself included - would do a particularly good job at that if we weren't allowed to use math and standards to lay the groundwork for properly scaled mechs.

[ScarecrowES]

oldradagast, on 06 May 2016 - 10:35 AM, said:

Oh, I agree on those points. Please don't think I am against volume normalization - I am not. I also agree that the Stalker's high hardpoints make a HUGE different. If the Awesome's first torso energy weapon slots were high mounted, they would probably at least be "somewhat competitive" in their current form.

All I'm getting at is that volume normalization, while good, will probably not eliminate the need for quirks to buff up mechs that are, in a typical situation, presenting more area to receive hits than other mechs. I hope PGI understands that - normalizing all mech volumes and then getting rid of all defensive quirks is not the answer, although the volume normalization is the right first step.

Normalization will absolutely NOT eliminate the need for quirks, you're absolutely right. Some mechs, based on their geometry alone, will just be better than other mechs. Unfortunately, as PGI found out trying to create an "algorithm" to do it, what makes a mech good is impossible to quantify. The survivability of a mech is related to a lot of factors that are entirely subjective. You can't really account for that in the geo unless you make every mech a simple object, or make them a 2d sprite.

So quirks are always going to be needed to balance out the intangibles. That's a given. But, if you want to produce the minimum omount of quirks that only factor in to mechs that absolutely need them, you're going to have to start with a purely objective baseline first.

The Stalker vs Awesome analogy is worth visiting here. Let's make clear right now, two mechs of the same tonnage, after the rescale, will not be the same "size" even if their tonnages are normalized. Shape has a lot of influence here. For instance, the Stalker has a very surface area-efficient shape compared to the Awesome. It's volume is more concentrated. You stand those two mechs side-by-side after the rescale, and that Stalker is going to look a lot smaller than the Awesome viewed from the front. It's certainly going to look bigger than the Awesome from the side though... it's going to have a LOT more surface area to shoot at from any angle but dead ahead, and it's going to have those arms to protect the torso like the Awesome has.

So it's going to be a trade-off there. Smaller and harder to hit from the front, but larger and easier to hit from the side, top, or bottom... with almost no ability to shield the exposed torso. On the other hand, supreme ability to shield the opposing torso. It's something people seem to be struggling with - you can actually view mechs from more than just the frontal view.

[ScarecrowES]

Zibmo, on 06 May 2016 - 08:11 AM, said:

And you know this because you've seen the math?

Having the Blackjack clearly undersized compared to its peers can mean several things: 1) its peers are enormous 2) it's way too small for a 45 ton mech 3) its peers are enormous AND it's too small for its weight class.

Saying "da maths" doesn't mean squat without a reference point. If the reference point is "the Atlas is perfect", then when comparing a 100 ton mech to a 45 ton mech, there should be a significant size difference. Unless, of course (like most of the IS mediums) their gigantic armor encases mostly air.

We already have several reference points, based on what PGI has said... for instance, there are certain models within a few percent of their real tonnage already in the game... the Centurion and Shadowhawk, for instance, are less than 2% off. You probably wouldn't be able to notice a 2% difference in size. So you can easily use those two mechs as a reference point. The current Nova model in-game is about 62 tons according to PGI. So you can get an idea of what 60-65 tonners will look like. In the 45-ton category, it seems like PGI is splitting the difference between the Vindicator and Blackjack. We don't know how far to the middle each one will go, but you can look at those two models and see that you're going to end up somewhere in between.

[ScarecrowES]

Ugh... sorry for all the consecutive posts folks... replying to multiple comments and the forum is not combining my posts as it usually would.

Zibmo, on 06 May 2016 - 08:20 AM, said:

What? This analogy is so out there I don't even know where to start. BJ (1 & 3) have been a couple of my favorite mechs since I started. Mine go 93kph. I do NOT expect them to be interior linemen. I do, however, expect to be able to play them. And that means taking a punch. Maybe not a haymaker like Foreman could. Maybe not a nuke. But a punch.

I was using the analogy given, and it very clearly works. The Blackjack weighs 45 tons. That means it has significantly less armor and internals than a 55-ton mech, which is the heaviest in the medium class. The Blackjack has a total possible armor value of 306 to the Shadowhawk's 370. The Shadowhawk has 120% of the armor of the Blackjack, at only 10 tons more. The Blackjack is not going to be able to take the hits that the Shadowhawk can take. Or at least it shouldn't.

If you don't have the ability to take a hit like the heavier mechs in your class, then you need to make up for that in other ways. You either forgo some weapons and pile on speed so you can get out of trouble, or you accept that you're fragile and pile on the guns, hoping to maximize the damage to the enemy before he takes you out. The Blackjack goes with the whole "slow and heavily-armed" choice. But it isn't being punished for it by being a glass cannon, as it should be... With quirks sitting where they are NOW, and with its extraordinarily small size, the Blackjack is probably the most survivable IS medium.

So the Blackjack... slow, can't take a hit, but packs a punch. This is going to be the kicker on the team. But you want to use it as a lineman against other linemen. By all rights, it should be getting its head knocked off. It's kinda not. That's a bad thing.

Edited by ScarecrowES, 06 May 2016 - 11:28 AM.

[oldradagast]

ScarecrowES, on 06 May 2016 - 11:07 AM, said:

So quirks are always going to be needed to balance out the intangibles. That's a given. But, if you want to produce the minimum omount of quirks that only factor in to mechs that absolutely need them, you're going to have to start with a purely objective baseline first.

Agree completely. My concern is not with the theory but the implementation after the mechs are normalized. PGI has a habit of "sort of" finishing tasks, with "Eh... that looks close enough." My fear is they'll normalize everything and then "get around to" adjusting quirks accordingly in a few months... years... when the Pinpoint skill does something... etc.

[mogs01gt]

RampancyTW, on 06 May 2016 - 10:12 AM, said:

MWO is a 3D game. Therefore its in-game objects have in-game volume. Balanced in-game volume is the best way to balance all possible profiles of a mech.I know you think you're being clever, but I can assure you, you're not.

clever,,no. Im trying to help individuals, like yourself, who obviously know nothing about balancing an FPS. In-game volume means nothing, what matters is the amount of pixels that are available to be aimed at.

Edited by mogs01gt, 06 May 2016 - 12:10 PM.

[operatorZ]

mogs01gt, on 06 May 2016 - 12:10 PM, said:

clever,,no. Im trying to help individuals, like yourself, who obviously know nothing about balancing an FPS. In-game volume means nothing, what matters is the amount of pixels that are available to be aimed at.

pixels that represent surface area, that represent volume, which represents mass.....

[DAYLEET]

Im starting to think the people who left their TBR when it wasnt godly anymore moved to the BJ1. Maybe we are heading to a game where the pilot being good is more important than the mech being good and that's called balance.

[Rampancy]

mogs01gt, on 06 May 2016 - 12:10 PM, said:

clever,,no. Im trying to help individuals, like yourself, who obviously know nothing about balancing an FPS. In-game volume means nothing, what matters is the amount of pixels that are available to be aimed at.

And because players aren't forced to stare directly at their opponents the whole game, profile (aka number of pixels) from EVERY vantage point matters. And the best way to get consistent, logical sizing, which will be reflected in overall, every-vantage-point profile, across a wide variety of mech shapes is volume.

Some shapes/hitboxes will be better at different things than others, but there at least won't be any mechs outright better due to size alone. Which is a great starting point.

I really don't get the whole front-profile fixation. The Banshee is massive from the front, but its bulky arms and slim side profile make it one of the tankiest mechs in the game.

Edited by RampancyTW, 06 May 2016 - 01:30 PM.

[Wintersdark]

oldradagast, on 06 May 2016 - 08:02 AM, said:

The problem with purely looking at volume alone is comparisons of mechs like the Stalker and the Awesome. I'd be willing to bet that if you look at their total volume - their appearance, effectively, from all angles - they are probably not that far off in scale from each other and where they are supposed to be. Except for one small problem - have a large front profile is MUCH worse than having a large side profile, which is why the Awesome falls apart if you look at it while the Stalker just keeps on going. Not all volumes are created equal.

Yeah, it's true that some mechs have profiles that hurt them. Not all mechs are created equal, as we all well know. Much like how a Misery has way, way better hardpoint options than an AWS-8V.

Balance work needs to be done beyond just scaling.

And this is also why there are some mechs receiving additional remodelling work too - the AWS in particular is seeing some more modelling.


I'll repeat yet again: The alternative to scale by volume is scale by Paul. Asking for custom scaling to address balance issues resulting from unfortunate geometry is asking PGI to make subjective comparisons of the mechs and then estimate the correct amount to scale those mechs to compensate for that subjective difference, as opposed to a mathematically derived objective scaling system being used currently.

This scaling by volume isn't a magic balance solution, but it's way, way better for the game overall than scaling by Paul.

[Wintersdark]

oldradagast, on 06 May 2016 - 10:51 AM, said:

Agreed. Math is the safest place to start as a baseline for proper mech scaling. Anything else, as you said, results in "balance by Paul / eyeballing / guessing / gut feeling." Not only has PGI demonstrated a poor track record on that, I'll be honest and say I have serious doubts much of anyone - myself included - would do a particularly good job at that if we weren't allowed to use math and standards to lay the groundwork for properly scaled mechs.

Exactly.

---

Sadly, this is why this discussion is so hard to have.

Scaling by balance and practical profile would be, in theory, far superior and have way better results. Sadly, it requires too much subjective guesswork.

So, those arguing for profile and such in this thread - they're all smart folks, and they make really great points, they're not wrong. But it's not a good option, because even if you give PGI the benefit of the doubt, scaling all the mechs subjectively like this would be a herculean task, and one doomed to failure from the start.

[ScarecrowES]

Wintersdark, on 06 May 2016 - 01:37 PM, said:

Yeah, it's true that some mechs have profiles that hurt them. Not all mechs are created equal, as we all well know. Much like how a Misery has way, way better hardpoint options than an AWS-8V.

Balance work needs to be done beyond just scaling.

And this is also why there are some mechs receiving additional remodelling work too - the AWS in particular is seeing some more modelling.


I'll repeat yet again: The alternative to scale by volume is scale by Paul. Asking for custom scaling to address balance issues resulting from unfortunate geometry is asking PGI to make subjective comparisons of the mechs and then estimate the correct amount to scale those mechs to compensate for that subjective difference, as opposed to a mathematically derived objective scaling system being used currently.

This scaling by volume isn't a magic balance solution, but it's way, way better for the game overall than scaling by Paul.

And this really is the whole point. Up until now, PGI just made mechs whatever size they felt looked right at the time. It's the reason you can put two mechs side by side that are the same tonnage and same body plan, but radically different in size.

The Vindicator and Blackjack are good examples here. There are a lot of intangibles thst might combine to determine why a mech might be good or bad to players. There are many variables. And as baffling as it is, size has always been a factor. It's difficult to balance intangibles, so we need to turn as many of those into real quantifiable constant as we can.

Would the Vindicator, for instance, be suitably survivable at the size of the Blackjack to overcome some of its other negative factors? I bet it would help.

[Satan n stuff]

ScarecrowES, on 06 May 2016 - 10:44 AM, said:

It's actually NOT literally impossible to quantify coastline length. It's a linear distance measurement. You start at one end, and then just measure to the other. There are hundreds of possible ways to accomplish this.

However, there is NO mathematical system that will allow you to estimate the relative size of an object based on an infinite number of silhouettes of that object. Such a system does not exist. Even if it DID exist, the most such a system could give you would equate to is a value representing a simplified estimate of the total surface area of an object, excluding surface areas that do not contribute to the outer profile of the object. So effectively, you've found a very contrived way to find surface area, which is something that's already available to you in the model source.

And since math tells us there is direct correlation between the surface area of an object and its volume, and given that your "profile scan" excludes any surface area of the object that will not contribute directly to volume, not only will you end up with the same relative scale regardless of which of the 3 methods you used - the profile estimate, surface area, or volume - but the ratio you came up with in the profile scan will actually sit much closer to volume than it does surface area. Regardless, volume will end up being the reference that is both the most accurate and the one with the least possible deviation from standards caused by erroneous surfaces.

So like I said... over and over... it doesn't really matter what you use. If you apply a standard which is based in actual math and takes the entirety of the model into consideration to all mechs - no matter which standard you use, the relative sizes of all mechs will end up the same. This mech will always be larger than that mech, and that mech will always be smaller than this mech, and so on. The ABSOLUTE sizes of those mechs may vary depending on the method used and the starting reference point, but the relative sizes will be the same no matter what.

The only way you get out of the Blackjack getting bigger is if you eyeball every mech... you don't apply a standard and look at mechs subjectively. So basically balance by Paul.

All of the methods you conveniently forgot to mention give a different result. I'm not sure if you're aware, but coastlines aren't straight and they don't just roll out a giant tape measure to get the right number, they don"t even measure the same at different times of the day. Coastlines are measured point to point and that will produce completely different numbers depending on how many points you use to measure and exactly where you place them, making it the perfect comparison to my proposed measurement, so try again.

As far as your pseudo-mathematical drivel goes, do you really believe that the numbers would resolve the same as total surface area, when half of most mech's features don't even contribute to the profile unless they happen to be sticking out to a side when you're looking at them? Those little bits tacked on to mech torsos like arms, guns, legs, they all have very high surface areas compared to the whole thing especially when looking at the combined total, but they add very little to the profile, only about one and a half leg and and a fraction of everything else on average.
Please come up with a mathematical proof if you think there's a direct correlation between total surface area and profile, otherwise stop spouting nonsense. I'll give you one hint, it doesn't exist because there is no such thing.

About the Blackjack specifically, I don't particularly care if it gets bigger or smaller, I have hundreds of mechs so I'm not going to worry that one of them might be a little weaker. What I am worried about is PGI rescaling all the mechs in a way that might make entire weight classes useless and will completely screw over chassis whose size are the only thing they have going for them. But of course being a mathematical genius as you are, surely you already saw the big picture, right?

Edited by Satan n stuff, 06 May 2016 - 01:56 PM.

[Davers]

RampancyTW, on 06 May 2016 - 09:46 AM, said:

No, they picked a volume/tonnage ratio that involves scaling down several mechs and scaling up others. Many lights are, in fact, getting smaller.

Threads like this giant crap fest are precisely why we don't see devs around very often.

Because the devs would rather have players argue amongst themselves than explain it themselves? Sounds like a great plan. They already make the players do their own Town Hall highlights. Maybe they can get them to data mine the game every month and make their own patch notes too?

[Rampancy]

Wintersdark, on 06 May 2016 - 01:46 PM, said:

Exactly.

---

Sadly, this is why this discussion is so hard to have.

Scaling by balance and practical profile would be, in theory, far superior and have way better results. Sadly, it requires too much subjective guesswork.

So, those arguing for profile and such in this thread - they're all smart folks, and they make really great points, they're not wrong. But it's not a good option, because even if you give PGI the benefit of the doubt, scaling all the mechs subjectively like this would be a herculean task, and one doomed to failure from the start.

The biggest issue with the "scale by profile" notion is there is no good way of defining what a mech's "practical profile" is. Because it's utterly situational. There's no way to develop balanced profile scaling that won't inevitably closely resemble a volumetric approach. So why not just remove subjectivity entirely?

I realize that you agree here, but the people you're calling "smart" are literally calling PGI incompetent and worse for using the best objective approach that accomplishes their desire for balanced profiles, all because it means slightly scaling up one mech that is unanimously considered the best IS mech in its weight class.

[Wintersdark]

Davers, on 06 May 2016 - 01:56 PM, said:

Because the devs would rather have players argue amongst themselves than explain it themselves? Sounds like a great plan. They already make the players do their own Town Hall highlights. Maybe they can get them to data mine the game every month and make their own patch notes too?

Right or wrong, because they used to explain things themselves, then they'd get drawn into these absurd arguments with dozens of players making stupid arguments and railing on and on, so they just stopped. Dev's explaining things (as much as I wish they would) ends up being a full time job on its own. The only case where it worked out well was in Karl Berg's thread, and even then he put in long hours on his own time to do it; but he wasn't discussing anything that was... subjective in nature, i guess.

[Davers]

Wintersdark, on 06 May 2016 - 02:09 PM, said:

Right or wrong, because they used to explain things themselves, then they'd get drawn into these absurd arguments with dozens of players making stupid arguments and railing on and on, so they just stopped. Dev's explaining things (as much as I wish they would) ends up being a full time job on its own. The only case where it worked out well was in Karl Berg's thread, and even then he put in long hours on his own time to do it; but he wasn't discussing anything that was... subjective in nature, i guess.

And yet other games famous for their toxic communities have Dev forum interaction, ESPECIALLY when they are reworking paid for content. Usually ongoing, starting several months before as well.

[keith]

ScarecrowES, on 06 May 2016 - 11:11 AM, said:

We already have several reference points, based on what PGI has said... for instance, there are certain models within a few percent of their real tonnage already in the game... the Centurion and Shadowhawk, for instance, are less than 2% off. You probably wouldn't be able to notice a 2% difference in size. So you can easily use those two mechs as a reference point. The current Nova model in-game is about 62 tons according to PGI. So you can get an idea of what 60-65 tonners will look like. In the 45-ton category, it seems like PGI is splitting the difference between the Vindicator and Blackjack. We don't know how far to the middle each one will go, but you can look at those two models and see that you're going to end up somewhere in between.

i'm sure its been mentioned in this thread but the nova won't be a "good" idea of what a 62 mech will looks like. now i'm done some 3d drafting, a bit different then modeling, alot more math involved. for stuff like this u would have to look at the surface area of the mech(front) and volume it takes up. take both of those facts and use them. the SA is what we would typical shot at, unless u twist. sadly as much math u put into the best way to balance this with out probley getting into higher end maths is to go by feel. but hey i've been done with this game for a while, good luck convincing pgi they done wrong

[Wintersdark]

keith, on 06 May 2016 - 03:24 PM, said:

i'm sure its been mentioned in this thread but the nova won't be a "good" idea of what a 62 mech will looks like. now i'm done some 3d drafting, a bit different then modeling, alot more math involved. for stuff like this u would have to look at the surface area of the mech(front) and volume it takes up. take both of those facts and use them. the SA is what we would typical shot at, unless u twist. sadly as much math u put into the best way to balance this with out probley getting into higher end maths is to go by feel. but hey i've been done with this game for a while, good luck convincing pgi they done wrong

You don't need to go into high end math; the 3D models have readily obtainable exact volume measurements. While complex math may be involved in determining that volume, it's fully in the background even from PGI's perspective.

Thus, it's a known factor that's easily obtainable, reliable and exact.

"Going by feel" is NOT a good way to go, not with PGI trying to scale 65 unique mech types in game right now and all those up coming chassis as well - particularly not the upcoming chassis that don't even exist yet. As a Founder, you've been around long enough to know that PGI "going by feel" to accurately scale that many mechs is an utter joke.

Instead, we can just use tonnage and volume, both factors that are exact and known.

[Alex Morgaine]

oldradagast, on 06 May 2016 - 08:02 AM, said:

Long story short, I expect once this resizing is done, a few real stinker mechs that were laughably oversized will suddenly be playable, a huge number will be made "technically" the right size, but will still be useless because of their shape (both hitboxes and hardpoint placement), and at least one mech will be rendered much weaker after being made larger so it is "technically" the correct size. We will then see a "never forget" thread on that mech and nothing will be done to fix it for years just like the laughable Catapult and its VCR ears.

True but the hopeful in me is that they can properly use standardized Mechs to address true size issues, and either Allow larger engine maxes or maybe even adjust problem mechs in size based off how bad they get hit box nerfed. At least the "x is to big wahhhhhhh" crowd will shut up with everything built to a standard... Or whine more, i dunno. Wish they incentivised the pts to get more accurate feed back instead of "waaahhhh i didn't DL pts but it sucks i know it does cause im a military programmer roboticist" that trends to happen. Although quirk pts 1.0 was hilarious..