378 replies to this topic

[Jin Ma]

7ynx, on 22 April 2014 - 11:47 PM, said:

The error in the assumptions of this thread are that the mass/volume densities are uniform across all mechs. Since this is SciFi, there is no authority on it and PGI can do what they want. That being said, there are obvious inconsistencies, the Quickdraw is one, most lights look much smaller than 1/3 to 1/5 the weight of the Atlas (using Atlas as the 100 tons standard).

I voted yes, but not for the reasons argued.

The argument in the thread was against people making volume arguments. under the assumption of equal density.

So unless you are one of those people assuming that volume is consistent across each of the mechs in respect to tonnage (which as the OP proved it was not). then the argument wasn't for you

Edited by Jin Ma, 26 April 2014 - 05:28 PM.

[Kaldor]

Lets lay this out simply:

First of all, this is not TT where someone in this thread basically said "on paper they are all the same"

Now we are playing what boils down to a FPS.
Size of a mech determines its ease of being hit
Smaller mechs are harder to hit. Smaller mechs are also generally faster.
Large mechs are easier to hit. Large mechs are also generally slower.

What does that tell us?
Size, speed, and armor all contribute to how long a mech lives without bringing up pilot skill.
Speed and armor is about in line for what the mechs should be.

This leaves the final variable that can be adjusted: size
If the Atlas is going to be the tallest mech in the game, then it should be the absolute upper limit in size vs armor. The Locust should be at the bottom of this scale. All mechs should fall in between.

Lets do some quick comparisons:

(Please bear in mind I used plugged values for the heights as I dont know exact values. If you can supply exact heights, please link them)

The Atlas can carry 614 points of armor and is our full sized model at about 20 meters tall which essentially has 30.7 points of armor per meter of height.

The Locust can carry 138 points of armor and is smallest sized model at about 5 meters tall which essentially has 27.6 points of armor per meter of height.

The Atlas having a higher armor rating per meter of height is good. Its a walking "shoot me" sign. The Locust having a slightly lower rating is good too as its tiny and can move at 170 kph.

Lets do a direct comparison:
The Quikdraw can carry 402 points of armor, which is roughly 2/3 the amount of the Atlas.
However, it is probably closer to 7/8 the size of the Atlas. This means that in spite its decent speed (100ish kph) and jump jets its very vulnerable and easy to kill.

You can run these comparisons all day....

What I am not saying is that all mechs should be cookie cutter in size. Using the upper and lower (5-20m in my above comparison) that should be where you start. Want to make a short fat medium? Fine. What to make a tall lanky light? Fine. However armor vs mech size vs speed is the ultimate formula and is directly responsible for how long a mech can live.

[DaZur]

I'd give my left testicle for my 60 ton Quickdraw to be only 10% larger than the 50 ton Centurion... Sadly, it's barely 5% smaller than a 100 ton Atlas.

I'd give my right one too but I've grown rather accustom to it's presence...

[Kaldor]

DaZur, on 27 April 2014 - 07:49 PM, said:

I'd give my left testicle for my 60 ton Quickdraw to be only 10% larger than the 50 ton Centurion... Sadly, it's barely 5% smaller than a 100 ton Atlas. I'd give my right one too but I've grown rather accustom to it's presence...

I dont know If Id give anything like that to PGI, but then again they may need them....

The QD is one of those mechs I want to really like. I have played all 3 to Master, but they are just so big that unless you never make a mistake, figure on getting shot up. Ive tried the LRM loadouts on them, its simply OK. Ive had decent success as a light hunter with SSRMs. SRM builds are a crap shoot as they are too hit or miss. I played mine like a bigger faster Hunchy SP with JJs which was fairly successful I guess.

Maybe when they get SRM hit detection working better I will re-visit them.

[Yanlowen Cage]

Don't forget the kintaro. It is huge. Also in tt the battlemaster is taller, but skinner then the atlas. I believe. As is the banshee.

[DaZur]

Kaldor, on 28 April 2014 - 06:22 AM, said:

I dont know If Id give anything like that to PGI, but then again they may need them....

The QD is one of those mechs I want to really like. I have played all 3 to Master, but they are just so big that unless you never make a mistake, figure on getting shot up. Ive tried the LRM loadouts on them, its simply OK. Ive had decent success as a light hunter with SSRMs. SRM builds are a crap shoot as they are too hit or miss. I played mine like a bigger faster Hunchy SP with JJs which was fairly successful I guess.

Maybe when they get SRM hit detection working better I will re-visit them.

I've been running my -5K as an effective flip of the bird to the "meta" with 4 x lLas + a single Ssrm pretty effectively. You are however 100% correct that it's completely unforgiving and require near flawless piloting to even shade toward being viable...

I'm 100% convinced if the Quickdraw was roughly 10% larger than a Centurion it would be pretty respectable...

I get that it's a Heavy class mech... But at 60 tons it's narrowly two duck farts away from being a Medium.

Edited by DaZur, 28 April 2014 - 08:55 PM.

[Adamski]

Hopefully now that they have finished pushing out a bunch of clan mechs, they can revisit some of the size issues that they have let languish.

If nothing else, the least they should do is provide the players with the numbers on why the sizes are currently set at their values.

[Strum Wealh]

Adamski, on 12 June 2014 - 03:48 PM, said:

Hopefully now that they have finished pushing out a bunch of clan mechs, they can revisit some of the size issues that they have let languish.

If nothing else, the least they should do is provide the players with the numbers on why the sizes are currently set at their values.

Or, we the players could take the time to figure it out on our own...

Strum Wealh, on 12 October 2013 - 05:00 PM, said:

Indeed - though, the argument is over which 'Mechs actually need to be re-scaled.

Here is the famous chart, courtesy of "Adridos" and "Bishop Steiner" circa January 2013:


By comparison, here is a table I made based on the information presented in some of my previous posts:

This table assumes a constant overall density, such that a decrease in volume produces a proportional decrease in mass.
This table also assumes that the Square-Cube Law is applied (such that a linear change in size/scale produces a cubic change in volume).
This table uses the MWO rendition of the Atlas (height determined to be 17.6 meters, as demonstrated on the Adridos/Bishop chart) as the reference point.
Note that this table is most applicable to upright (e.g. the torso is, or is nearly, "taller" than it is "wide" or "deep"/"long") humanoids (such as the Commando, Centurion, and Atlas); it makes no attempt to be applicable to "hunched-over" 'Mechs whose torso is significantly "wider" or "deeper/longer" than it is "tall" (such as the Raven, Dragon, Catapult, or Stalker). Nor does it attempt to address the differences in build between upright humanoid 'Mechs (e.g. the Centurion and Trebuchet being "skinnier"/"lankier" than the Hunchback, or the Atlas).

What the table does do, however, is show how tall an Atlas would be if a rescaled to the mass of any other given BattleMech, assuming overall density (that is, total mass divided by total volume) is held constant and the Square-Cube Law is applied (as a number of proponents of 'Mech rescaling have suggested).

Now, for a comparison between the Adridos/Bishop chart and my previous table:

Of note is how each of the upright 'Mechs except the Commando is within one meter (or, alternatively, within a single-digit percentage) of the height suggested by "scaling to the Atlas".

This data supports three statements:

• With regard to the upright humanoids presented in the Adridos/Bishop chart, the Medium 'Mechs (and the Heavies, and the Assaults) are more-or-less "the right size", or close enough that the differences are negligable.
  
• With regard to the upright humanoids, the Light 'Mechs (certainly the Commando, and probably the Spider as well) are "too small".
  
• One method of "improving gameplay balance" is to bring the Light 'Mechs in-line with the other weight classes with regard to scaling, and having all subsequent 'Mechs adhere to the same scaling.

The Commando & the Spider (and likely the Locust, Jenner, and Raven as well) are in far greater need of a re-scaling (specifically, one that results in an up-sizing) than the Centurion & the similarly-sized Trebuchet.

Strum Wealh, on 13 October 2013 - 02:12 PM, said:

Though, the questions to be asked become "what are our reference points?" and "how much variation from 'in-line' is acceptable?", especially when a number of 'Mechs are defined by their geometry (e.g. the Centurion & Trebuchet versus the Hunchback; the former are tall and lean while the latter is short and stout).

Which comes back to the point of density (that is, the ratio of the 'Mech's total mass to the total volume of the in-game model) as a factor in 'Mech scaling.
As Bishop pointed out, "Military hardware almost always is designed as compact as possible, to minimize target profile, and increase it's versatility. Just as one doesn't tend to see wildly disparate sizes amongst MBTs of today, you will see only minor fluctuations in density and such in military hardware".
Given this rationale, the fact that most 'Mechs are made of the same materials, the realities of the Square-Cube Law ("When an object undergoes a proportional increase in size, its new volume is proportional to the cube of the multiplier and its new surface area is proportional to the square of the multiplier."), and the use of the MWO Atlas as a reference point, I created the following table:


At its most literal, the table shows that "a MWO Atlas, assuming the Square-Cube Law is held to be true and assuming that density is held constant would be of the height given in the table if its total mass (and, given constant density, its total volume) were reduced to that of any given weight bracket".
More broadly, the table shows what is arguably the appropriate approximate height of any upright humanoid MWO 'Mech of the same density as (and, ideally, approximately similar in build to) a MWO Atlas.

(Incidentally, it also generally works given TT values, as well; if a TT Atlas is ~14.00 meters tall, a 20-ton upright humanoid like the Thorn should be ~8.19 meters tall and a TT Commando (25 tons) or a Mongoose should be on the order of ~8.82 meters tall.)

Using the values from Adridos'/Bishop's chart (circa Jan. 2013) produces the following:

With the few numbers I do readily have, those 'Mechs that are relatively "upright" (as opposed to "hunched-over" like the Raven, Dragon, Catapult, and Stalker - where the torso/body is significantly "longer" than it is "tall") other than the Commando are less than one meter off of where they would be predicted to be (or, alternatively, they are within single-digit percentages (and generally close enough to fall within or below some of the common thresholds of statistical significance) of where they would be predicted to be).

Without knowing the actual height in meters, my model would predict the Quickdraw (as a 60-ton upright humanoid) to be on the order of ~14.84 meters tall, with a ~5% variation allowing for heights up to ~15.58 meters to still fall within the realm of "arguably acceptable".
Likewise, the Shadow Hawk (as a 55-ton upright humanoid) would be predicted to be on the order of ~14.42 meters tall (or up to ~15.14 meters, given a ~5% variation), the Thunderbolt (as a 65-ton upright humanoid) would be predicted to be on the order of ~15.25 meters tall (or up to ~16.01 meters, given a ~5% variation), and the BattleMaster (as an 85-ton upright humanoid) would be predicted to be on the order of ~16.67 meters tall (or up to ~17.50 meters, given a ~5% variation).
It would be interesting to see the actual heights of the models, to see how close the above actually gets...

(Then again, the second table also shows that the Commando is nearly twice as far off, in terms of actual measurement (that is, meters) as well as by percentage, than any of the listed Mediums, Heavies, and Assaults. I would be unsurprised to learn that the Spider is also off from where it "should be "by a rather large degree, nor would I be surprised to learn that the Lights were intentionally made to be "too small", making them into outliers in order to enhance their advantages versus the other classes... )

[STEF_]

When I saw the centurion first time, I simply do not recognize it... tooooooo large!
Moreover the arm position and mechanics are completely different
"True" centurion is not this one in mwo.

Edited by Stefka Kerensky, 25 July 2014 - 11:21 PM.

[McTough]

Even if the catapult is too big, if shrunk wouldn't the next thread be about catapults being OP because they're too hard to hit in their natural environment...at long range?

Edited by McTough, 11 August 2014 - 04:33 PM.

[wolf74]

Strum Wealh, on 25 July 2014 - 04:47 AM, said:

spin

The Problem with just using a Height you misses Y & Z Factors. The Best way to scale them would be Volume to Tonnage Ratio.

AKA Load the Model in to a program that can give you it's Volume.

(The following is using MADE UP NUMBERS Due to the Fact I do not have said Program to give me a model Volume)

100ton Atlas = Volume in 65m³ which would give us a 1ton = 0.65m³
Now you can scale all the other model so that their tonnage has the same ratio of 1:0.65 with a +- of 0.01 error factor.

Edited by wolf74, 13 August 2014 - 11:02 AM.

[Strum Wealh]

wolf74, on 13 August 2014 - 11:00 AM, said:

The Problem with just using a Height you misses Y & Z Factors. The Best way to scale them would be Volume to Tonnage Ratio.

AKA Load the Model in to a program that can give you it's Volume.

(The following is using MADE UP NUMBERS Due to the Fact I do not have said Program to give me a model Volume)

100ton Atlas = Volume in 65m³ which would give us a 1ton = 0.65m³
Now you can scale all the other model so that their tonnage has the same ratio of 1:0.65 with a +- of 0.01 error factor.

"Though, the questions to be asked become "what are our reference points?" and "how much variation from 'in-line' is acceptable?", especially when a number of 'Mechs are defined by their geometry (e.g. the Centurion & Trebuchet versus the Hunchback; the former are tall and lean while the latter is short and stout)."

"Which comes back to the point of density (that is, the ratio of the 'Mech's total mass to the total volume of the in-game model) as a factor in 'Mech scaling."

"This table assumes a constant overall density, such that a decrease in volume produces a proportional decrease in mass.
This table also assumes that the Square-Cube Law is applied (such that a linear change in size/scale produces a cubic change in volume)."

"At its most literal, the table shows that 'a MWO Atlas, assuming the Square-Cube Law is held to be true and assuming that density is held constant would be of the height given in the table if its total mass (and, given constant density, its total volume) were reduced to that of any given weight bracket'.
More broadly, the table shows what is arguably the appropriate approximate height of any upright humanoid MWO 'Mech of the same density as (and, ideally, approximately similar in build to) a MWO Atlas.
"

My posts already take into account such factors as volume & density (in fact, the Scaling factor is actually computed from volume & tonnage ratios, such that (for example) a 20-ton 'Mech should be roughly 58% of the size, in any dimension or combination of dimensions, of a 100-ton 'Mech of the same average density & roughly the same proportions; remember, volume increases cubically in response to linear changes in scale), and the multiple dimensions of scaling (such as limb radius/diameter vs limb length, torso width & depth vs torso height, and so on).

Edited by Strum Wealh, 13 August 2014 - 11:43 AM.

[Leopardo]

any way i want to re-scale the mechs and maps - i dont fell like im in big robot. scale between mechs i think normal

[s0hno]

I saw a Shadow Hawk and Storm Crow standing next to a Stalker lately - they both were taller than the 85 tons monster.
Thaaaat looked totally weird.

Well, if someone of my height weights 150% compared to me, he, ehm... probably has very heavy bones?
But assuming the Stalkers armor is not made of plumb, something definitely went wrong here.

PGI, do something, people here even already did the hard math :-)

Edited by s0hno, 05 December 2014 - 03:33 AM.

[Catalinasgrace]

I had seen in another thread where someone said that re-scaling was a no go... Is this confirmed? I certainly hope it is just some goofball on the forums talking out of his back side...

[Adamski]

Yep, Dennis from the Art Department has his head so far up his own ass, he thinks that mech scaling is in a good place, as seen here:

http://mwomercs.com/...vs-42-answered/

So abandon all hope ye who enter here, you will have to rely on quirks to fix your stupidly large mechs, because Dennis sure as hell wont help you.

[Adamski]

Oh, and regarding this silliness:

Quote

My posts already take into account such factors as volume & density (in fact, the Scaling factor is actually computed from volume & tonnage ratios, such that (for example) a 20-ton 'Mech should be roughly 58% of the size, in any dimension or combination of dimensions, of a 100-ton 'Mech of the same average density & roughly the same proportions; remember, volume increases cubically in response to linear changes in scale), and the multiple dimensions of scaling (such as limb radius/diameter vs limb length, torso width & depth vs torso height, and so on).

Volume has a 1:1 ratio with weight assuming the same materials. 1 litre of water (measure of volume) is 1 kilogram (measure of weight). This does not change no matter how many litres or how many kilograms.

A 20 ton mech would displace 20 kilo-litres, and a 100 ton mech would displace 100 kilo-litres.

[MadLibrarian]

Last I heard some mechs were still using the old non dynamic model geometry, so there still might be a geometry pass at some point. It's the same with maps, some are from a different era and look out of place in comparison with the new stuff. Here's hoping they use the time following CW's beta release for fixing all the old nagging problems, adding simple obvious features(like chat scrolling), and bringing old assets into line with the new.

I think the addition of Solaris could add to the importance of consistent mech scaling and mech balance within weight classes. Geometry and scaling might be a good thing for the art team to work on while the other guys design Solaris.

[Trashhead]

I really hope this will happen, but i think PGI has their hands full atm.
Besides, re-scaling those mechs might also involve re-scaling textures... and weapons.
So this is probably more work then we assume it is.

Apart from this, yes please, PGI.
It's always odd to see a Catapult next to a Stalker, knowing the later one is 20 tons heavier, but looks smaller.
(And, talking about the Stalker, please make the torso slightly bigger - maybe it's only me, but the Torso-size-to-leg-size-ratio looks a bit weird here.)

[Bishop Steiner]

justice bump