34 replies to this topic

[VirtualRiot]

[UPDATE: Added side pixel count for more accurate numbers]
For a long time now it has always occured to me that the best mechs out there always seem so... dense. That is, the density of the mech in terms of Volume / Tonn.

I decided to do a little experiment to satisfy my own curiosity and decided to share it here.

First I took one of the most popular mechs, the Dire Wolf. I took a screen cap of it in the lab, making is face frontmost to the camera as much as I could.




I then took the photo into paint and edited it so that I was only left with the mech





Then, using a pixel counting program, counted the number of pixels that were pure white. I then subtracted that amount from the total number of pixels, to come up with a number that I used as a representation of the mechs volume.

I then did the same thing with a mech that people tend to find very bad, the Quickdraw






Anyways, the result for the pixel measurements was thus;

Dire Wolf:
340,852 Pixels Front
253,605 Pixels Side
Total: 594,457 Pixels
Quickdraw:
280,424 Pixels Front
181,040 Pixels Side
Total: 461,464 Pixels

If we divide this number by the tonns of the mech we can find its Pixels Per Tonn (P/T) which is a representation of its volume, and thus how dense the mech is. Remember that in this case, the less pixels per tonn, the more dense the mech.

Dire Wolf: 5945 P/T
Quickdraw: 7691 P/T

We can see that the surface area per tonn of these mechs is off by quite a bit. The Quickdraw has ~22% more surface area per tonn than the Dire Wolf.

This is what they currently look like side by side


This is what they should look like


I hope to compare more mechs in the future when I have time to have a more accurate comparison between mechs, and compare that to how good people seem to think they are. Take from this what you will, good day.

Edited by VirtualRiot, 21 May 2015 - 07:47 PM.

[Wintersdark]

Now, turn those mechs sideways and do it again. Even if you don't want to do this in 3d, depth is critical too. The Direwolf is a massive block, while the Quickdraw is wide but a tiny fraction of the depth.

The quickdraw doesn't have even remotely as much volume as the Direwolf.

[Khobai]

It isnt exactly a secret that the mechs are scaled poorly...

You have mediums like the Trebuchet that are as tall as Awesomes.

Quote

Now, turn those mechs sideways and do it again. Even if you don't want to do this in 3d, depth is critical too. The Direwolf is a massive block, while the Quickdraw is wide but a tiny fraction of the depth.

The quickdraw doesn't have even remotely as much volume as the Direwolf.

Height is actually the worst dimensional attribute for a mech to have. Because height makes you easier to hit from every direction. Front, back, AND sides.

Thats why tall mediums like the Trebuchet just dont work.

Edited by Khobai, 21 May 2015 - 03:47 PM.

[Ghost_19Hz]

This is brought up so often and so many people agree on it that its astonishing this is not on PGI's "to do list".

[Neput Z34]

Survivability is a combination of armor values relative to the size or surface area of mech combined with maximum possible speed.
The first post illustrates the over all size / surface area to tonnage disparity, but it doesn't take into account the size of individual hit box and armor amount associated with it.

In my limited experience of driving QKD, most of the time it dies due to a loss of both legs, CT or combination of leg+arm(s)+CT. Very rarely it died to side torso and that was mostly due to "pilot error"
I don't own any Dire Wolfs, but those mostly seem to die to CT or both side torsos, from what I have seen.
Conclusion, the components listed above have; low armor to surface area ratio / low armor density / "weak armor" / poor hit box dimensions / etc,.

This takes scaling and "balance" to the next level of higher mathematics.

I wouldn't use "Volume" and "Density" to describe some of the mech models because those terms are mutually exclusive to whoever did the dynamic missile hard points to look like "tree cancer", increasing the size of a mech. The Catapult-A1 being the most glaring example, but I digress.

In my humble opinion*, this is were IGP/PGI "screwed the pooch" in the balance department by inconsistent mech sizes and their related hit boxes.
*Opinions are like a "certain part" of a human anatomy, everyone has one.

[VirtualRiot]

Wintersdark, on 21 May 2015 - 03:44 PM, said:

Now, turn those mechs sideways and do it again. Even if you don't want to do this in 3d, depth is critical too. The Direwolf is a massive block, while the Quickdraw is wide but a tiny fraction of the depth.

The quickdraw doesn't have even remotely as much volume as the Direwolf.

Got it coach, see OP

[Juodas Varnas]

Can you compare it to other Heavy mechs? And maybe more Humanoid Assaults? Because i think it would make for a better comparison.

I've been tweeting russ about quickdraw's size, movement archetype (it being large, and not medium like most IS heavies), but he never responded.

Edited by Juodas Varnas, 21 May 2015 - 08:13 PM.

[Mercules]

One other thing to take into consideration is tonnage left over for "gear" like weapons, armor, and equipment. Long ago I created a spreadsheet for Battletech and when I designed new mechs I found out the optimal tonnage for it's speed. If I wanted a mech that could go X speed I would then find the tonnage that would have the most left over after the tonnage needed to be devoted to internal structure, engine/gyro, and other fixed items like cockpit/scanners, and such.

As SPEED can be as important for survival as hitboxes(and speed combined with good hitboxes is golden) you may want to consider that as well. Maybe do a comparison of left over tons once the mech has a max sized engine both XL and Standard and then use that to measure against surface area.

[VirtualRiot]

Mercules, on 21 May 2015 - 08:20 PM, said:

One other thing to take into consideration is tonnage left over for "gear" like weapons, armor, and equipment. Long ago I created a spreadsheet for Battletech and when I designed new mechs I found out the optimal tonnage for it's speed. If I wanted a mech that could go X speed I would then find the tonnage that would have the most left over after the tonnage needed to be devoted to internal structure, engine/gyro, and other fixed items like cockpit/scanners, and such.

As SPEED can be as important for survival as hitboxes(and speed combined with good hitboxes is golden) you may want to consider that as well. Maybe do a comparison of left over tons once the mech has a max sized engine both XL and Standard and then use that to measure against surface area.

So you're saying make a graph to see if there is a correlation between max engine & pixels per tonn?

[William Mountbank]

Neput Z34, on 21 May 2015 - 04:35 PM, said:

Survivability is a combination of armor values relative to the size or surface area of mech combined with maximum possible speed.
The first post illustrates the over all size / surface area to tonnage disparity, but it doesn't take into account the size of individual hit box and armor amount associated with it.

What you say is true, but speed is balanced by engine tonnage and hit boxes by play style, whereas pixel count is hard baked into each mech, and currently has to be balanced with quirks. I think it's easier with these things to keep the initial examination simple - no# of pixels per tonne is easy, and puts real numbers on why people say the QKD is a poor mech. I happen to like the QKD and play it a lot, but that's because it's fun and a challenge, not because it's 'good'. Working in the mechlab can solve all other problems, but not how much pixel exposure a mech has in game.

[Mercules]

VirtualRiot, on 21 May 2015 - 08:58 PM, said:

So you're saying make a graph to see if there is a correlation between max engine & pixels per tonn?

No, more like a comparison between several things.

1. Max speed
2. Tons left over once Max speed/armor is achieved
3. Pixel size of mech.
4. Armor value


We might find out that something like the hunchback moves just enough, with enough of a weapon loadout, just enough armor, that it's pixel size is pretty dang good(this is a hunch). Light mechs have low tonnage left and low armor value, but probably have very good pixel ratio and with speed come out pretty good. Slow tall mediums probably not as well.

[Kiiyor]

SCIENCE! Great work.

I'd suggest using GIMP or Photoshop to delete the needful, far less time intensive than paint.

I did something ver similar with a handful of mechs a little while ago, and found that the Awesome was a bigger porker than the dire from the front.

If you were feeling particularly extravagant, you could break your area counts down to rag doll components :-)

[Navid A1]

MWO scales are pretty off on quite a number of mechs.

another example: (either Quickdraw is way too big or the zeus is way too small)

[Dawnstealer]

This has been a known issues since the Centurion rolled out (or the ever-present Stalker/Catapult comparison) - it's definitely been present for a while.

Lore-wise, they're pretty close to what they should be, but in a game like this? They need to tweak the lore a bit and scale them to their weight class. As a lore-junky, that's painful to say, but for the person that has to hop in some of these mechs and run around in them? Not so much.

[Deathlike]

We're never going to have properly scaled mechs.

[Wintersdark]

VirtualRiot, on 21 May 2015 - 07:27 PM, said:

Got it coach, see OP

yeah, I get it. My point was, you need at least a side profile as well for this to have any meaning at all. Just front on is completely worthless.

See: awesome is bigger than the Direwolf from the front... But there's what, a 20t difference and the Direwolf is easily 4 times as thick. If anything, the Awesome should be bigger than it is, if the Direwolf is right.

Aha! My apologies, I just realized you edited your OP. Disregard the previous post, I misunderstood your reply.

Edited by Wintersdark, 22 May 2015 - 09:22 AM.

[Strum Wealh]

I am also curious as to why you (VirtualRiot) used a comparison between a Daishi and a Quickdraw - two 'Mechs with vastly different body types - as opposed to a comparison between an Atlas and a Quickdraw.

As for the actual math, see this post.
Note that an Atlas scaled down from "100 tons" to "60 tons" results in a 40% decrease in overall volume, but only a 2.76-meter decrease in height (from 17.6 meters to 14.84 meters).
On top of that, an actual Quickdraw would (or should) have a leaner build than a "Quickdraw-sized Atlas", such that the former would end up "somewhat-thinner/skinnier-but-somewhat-taller" in comparison to the latter.

[AlexEss]

I find it endearing when people talk about mech scale...

Especially as no real scale has ever been established for many mechs and nobody ever followed the few guidelines that where.

Also i would nor trust the mechbay camera, it is jittery... (not to mention the cockpit view that is seven way of wonky on it´s own even before people start fiddeling with the FoV)

But a good start i guess. The proper way to do it ofc is some nice magic and a 3d modeling software...

[Shredhead]

M2 Bradley; 6.55m x 3.6m x 2.98m at 30 tons

M1A2 Abrams; 7.93m x 3.66m x 2.44m at 62 tons


Quick! Someone tell them to rescale their vehicles!

OMG!

Quickly now, before the sand people get it!

[TheCobra]

Shredhead, on 22 May 2015 - 12:49 PM, said:

M2 Bradley; 6.55m x 3.6m x 2.98m at 30 tons

M1A2 Abrams; 7.93m x 3.66m x 2.44m at 62 tons


Quick! Someone tell them to rescale their vehicles!

OMG!

Quickly now, before the sand people get it!

Comparing a MBT to an APC doesn't make sense. One is a troop transport with a lot of hollow space for, you know.... transporting troops. Your analogy is the same as comparing an atlas to a dropship.