81 replies to this topic

[zagibu]

Seeing Tennex's nice graph showing the mechs profiles in relation to each other made me recreate his data, in order to verify it, but also because I wanted to publish the underlying data as well.

Here is a zip with orthogonal screenshots of all current mechs that I used to calculate the profiles and an OO calc document containing all numbers both of Tennex and me: dropbox-link

The measurement method is not 100% exact of course, because in order to do this, we'd have to have the profiles of the actual hitboxes, not the 3d art. But since the hitboxes should follow the 3d art pretty closely, the trends found in this way should still be relevant.

The square root of the area of the profiles has been divided by the cubic root of the mech's tonnage. This yields the value PPTA, which tells us how much surface area the mech is exposing per ton and allows us to directly compare mech profiles.

Here is a picture showing the relative PPTA values of all mechs compared to the average PPTA value both for my own and Tennex's analysis:



As you can see, the values are pretty close.

So what does this data tell us? The first insight is that in general, light mechs are much too small and everything else is slightly too large, with a few extreme mechs being much too large. Remember, this "size" is normalized by tonnage, and in the ideal case, all mechs would have 100% here, which would mean they all expose the same profile per ton. This is clearly not the case.

It also shows us that some mechs, which feel much too large, actually don't derive that much from the average, like the Awesome, for example. This mech probably needs remodeling of the whole torso area instead of a simple proportional rescale.

We can also see that there is a reason so many players are raging about the light mechs. Some of the worst offenders are MUCH too small, and they also seem to cover the mechs that attract the most rage (Spider, Firestarter, Arctic Cheetah).

Another interesting insight is that the latest mechs all seem to be scaled smaller than mechs with comparable tonnage. The Mauler has the second smallest assault mech profile after the Zeus, and both the Black Knight and the Crab are much smaller than anything else in their tonnage.

If PGI ever starts with the rescaling project, they can take a look at the lower end of this list for a hint where to start. Catapult, Nova, Dragon, Quickdraw are all good candidates.

[edit]Karl Streiger has done some interesting work comparing the armor values (including quirks) of mechs with the mech profile data. Read all about it here. It yields yet a slightly differently ranked list of a "size" concept, basically dividing the size of the mech by its armor+structure, or calculating a sort of "density", if you wish, which determines how squishy a mech is normalized over hitpoints.

The winners are: Spider, Commando 1D and Panther 10P, which are all below 50 pixels per hitpoint.

Average mechs are around 75 PPHP, such as Kitfox, Jagermech and some Grasshopper variants.

The clear losers are: Maddog, Hellbringer and most Nova variants, which are much over 90 PPHP
.
What this means is if you take the HP of a Spider, divide it by the HP of a Maddog, then use this ratio to scale the Maddog down, the Spider will still have a mech profile that is HALF the size of the Maddog.
[/edit]

[edit]New data below[/edit]

Edited by zagibu, 23 May 2016 - 03:22 PM.

[Tennex]

oh nice, u got my black knight stuff lol, i never graphed it but i did calculate the profile for it.

Yeah BK is extreamly undersized because its so skinny. Wish they would stick to a consistent art style/girth for the mechs in each chassis to avoid the issue that BK and Grasshopper have

Here is updated graph



If you correlate the graph to visuals, its not suprising that the black knight is so close to the shadowhawk on the chart. Sadly there is a 20 ton difference.

Edited by Tennex, 10 October 2015 - 06:52 AM.

[Hotthedd]

I love people like you. Those graphs are awesome.

Although a complete re-scale is unlikely to happen, and even if it did, there would still be variation, one possible solution could be to allow 'mechs with more surface area per ton to have heat dissipation quirks (and institute heat penalties).

[Kiiyor]

zagibu, on 10 October 2015 - 06:02 AM, said:

Seeing Tennex's nice graph showing the mechs profiles in relation to each other made me recreate his data, in order to verify it, but also because I wanted to publish the underlying data as well.

Here is a zip with orthogonal screenshots of all current mechs that I used to calculate the profiles and an OO calc document containing all numbers both of Tennex and me: dropbox-link

The measurement method is not 100% exact of course, because in order to do this, we'd have to have the profiles of the actual hitboxes, not the 3d art. But since the hitboxes should follow the 3d art pretty closely, the trends found in this way should still be relevant.

The square root of the area of the profiles has been divided by the cubic root of the mech's tonnage. This yields the value PPTA, which tells us how much surface area the mech is exposing per ton and allows us to directly compare mech profiles.

Here is a picture showing the relative PPTA values of all mechs compared to the average PPTA value both for my own and Tennex's analysis:



As you can see, the values are pretty close.

So what does this data tell us? The first insight is that in general, light mechs are much too small and everything else is slightly too large, with a few extreme mechs being much too large. Remember, this "size" is normalized by tonnage, and in the ideal case, all mechs would have 100% here, which would mean they all expose the same profile per ton. This is clearly not the case.

It also shows us that some mechs, which feel much too large, actually don't derive that much from the average, like the Awesome, for example. This mech probably needs remodeling of the whole torso area instead of a simple proportional rescale.

We can also see that there is a reason so many players are raging about the light mechs. Some of the worst offenders are MUCH too small, and they also seem to cover the mechs that attract the most rage (Spider, Firestarter, Arctic Cheetah).

Another interesting insight is that the latest mechs all seem to be scaled smaller than mechs with comparable tonnage. The Mauler has the second smallest assault mech profile after the Zeus, and both the Black Knight and the Crab are much smaller than anything else in their tonnage.

If PGI ever starts with the rescaling project, they can take a look at the lower end of this list for a hint where to start. Catapult, Nova, Dragon, Quickdraw are all good candidates.

This, this is some fantastic science.

It gives a great indication of the fattest fatties, but we shouldn't forget proportions either. Hitboxes make a pretty huge difference to viability in addition to scale - some mechs, like the StormCrow, are a long way down the list, but don't suffer as much because the allocation and proportions of their hitboxes are excellent. Others, like the Awesome, are around the same area - yet the Awesome has midget legs and an enormous torso area. In fact, I recently measured the torso pixels only of a few fat mechs and found that the Awesome has a larger frontal torso profile than a Dire!

Great work. I shall ponder upon this.

[zagibu]

Yes, it would be interesting to do some image analysis and calculate eccentricity values to get a better feel of how "fat" a target really is.

[Kiiyor]

zagibu, on 10 October 2015 - 06:16 AM, said:

Yes, it would be interesting to do some image analysis and calculate eccentricity values to get a better feel of how "fat" a target really is.

I've always wanted to start with separate calculations for the legs and torsos, and maybe work my way down to individual hitboxes. I started with some Assault mechs once, but got bored and wandered away after I realized geometry often changed a great deal on mechs like the MadCat depending on loadout.

[zagibu]

Kiiyor, on 10 October 2015 - 06:26 AM, said:

I've always wanted to start with separate calculations for the legs and torsos, and maybe work my way down to individual hitboxes. I started with some Assault mechs once, but got bored and wandered away after I realized geometry often changed a great deal on mechs like the MadCat depending on loadout.

That's true. This analysis only covers the best case, without weapons. Add missile boxes to the Mauler, and it should slip a few slots down the list. Same with other mechs that have detachable parts of significant size, e.g. Ebon Jaguar, Thunderbolt, etc.

[Sarlic]

Thanks!

Interesting and something i already knew but numbers says it all.

[zagibu]

Sarlic, on 10 October 2015 - 06:36 AM, said:

Thanks!

Interesting and something i already knew but numbers says it all.

I was surprised by a few, for example the Stalker being slightly over average.

[FupDup]

Given the current state of balancing, making lights and many mediums bigger would just make even more people flock to heavies and assaults.

Edited by FupDup, 10 October 2015 - 08:35 AM.

[zagibu]

FupDup, on 10 October 2015 - 08:33 AM, said:

Given the current state of balancing, making lights and many mediums bigger would just make even more people flock to heavies and assaults.

Of the medium mechs, only the Crab, Blackjack and Cicada are noticeably smaller than average. The others are either where they should be or larger.

Also, this is just a neutral presentation of the data. You are probably right that it isn't a good idea to make lights bigger. It might be better for balance to keep them underscaled. But if you then compare for example the Firestarter with the Kitfox, you see that there is a pretty big discrepancy.

[Sarlic]

Perfectly suits the line:

http://mwomercs.com/...l/page__st__180

Crabs are way too small. The Centurion looks a giant next to it.

Edited by Sarlic, 13 October 2015 - 06:34 AM.

[Dino Might]

I think Crabs are sized well. The Centurion is just too big. Most mediums need a size reduction. The Crab is darn tough to kill because of its size and shape. If more mediums were like it, they'd be more popular. As it is now, who really has a tough time with Shadowhawks, Griffins, Wolverines, etc? They're gigantic targets that don't move much faster than heavies.

[Erkki]

Yeah the Crab is too small compared to its weight... Now imagine if it had been given quirks similar to HBK or CN9's. It didn't receive laser duration quirks at all, just some structure points and small range, heat or cooldown quirks depending on the variant. And AMS range lol.

PGI didn't scale it right but that mistake at least helps it to not be DOA. People are likely to play it even months from now as often as most other mediums... Vindi and Kintaro are still sad.

[Sarlic]

Dino Might, on 13 October 2015 - 06:42 AM, said:

I think Crabs are sized well. The Centurion is just too big. Most mediums need a size reduction. The Crab is darn tough to kill because of its size and shape. If more mediums were like it, they'd be more popular. As it is now, who really has a tough time with Shadowhawks, Griffins, Wolverines, etc? They're gigantic targets that don't move much faster than heavies.

Or that. That would be more ideal.

But if i look realistic at the whole game. Newer mechs gets the win, older mechs gets the sack.

I mean, the rescale should have bee done a.s.a.p.

Now some mechs are still suffering the terrible scale format and not to mention bad hitboxes.

Edited by Sarlic, 13 October 2015 - 06:54 AM.

[QuulDrah]

in WWII the military used to hand out these nifty charts. designed for spotters to correctly identify and report incoming planes...
Example

could some of you photoshop wizards with too much free time come up with something similar?
one chart for IS, one for Clan?

[MeiSooHaityu]

Tennex, on 10 October 2015 - 06:06 AM, said:

Is that a new mech on the left? Cool!


Oh wait, nevermind I recognize it now, it is a Shadow Hawk. Man, it has been AGES since I seen one ingame, I forgot what they looked like.

It's amazing how they went from the hot stuff to MIA when the quirks first went live. Maybe it will rise again after the next quirk pass.

Oh, and yes, the Shadow Hawk is way too big. It probably should have been included in a rescale, but when you only want to pick a few, many deserving mechs get left out.

[Juodas Varnas]

MeiSooHaityu, on 13 October 2015 - 07:16 AM, said:

Oh, and yes, the Shadow Hawk is way too big. It probably should have been included in a rescale, but when you only want to pick a few, many deserving mechs get left out.

Well, it DID manage to get more votes than the Trebuchet.
So that's something... Something absolutely ridiculous, to be honest.

[zagibu]

Juodas Varnas, on 13 October 2015 - 07:30 AM, said:

Well, it DID manage to get more votes than the Trebuchet.
So that's something... Something absolutely ridiculous, to be honest.

In Tennex's chart, it's somehow bigger than the Trebuchet. Not sure why, though. In the profiles I made, it's smaller (although not a whole lot).

[Sarlic]

Suprisingly low comments.

The Crab is too tiny. Even a monkey can see it.

The question is why now and why didnt PGI decide to re-scale mechs who need it. One per month should be do-able.

It feels like the re-scale program is on a backburner while the new toys get the right attention.

The Centurion for example should have the approciate scale-down like our current scaled Crab.

Edited by Sarlic, 18 October 2015 - 01:05 PM.