Jump to content

Ever Wondered How Many Urbanmechs It Takes To Kill A Dire Wolf? (Mwo Loadout Simulator)


27 replies to this topic

#1 fat4eyes

    Member

  • PipPipPipPipPipPip
  • The People's Hero
  • 491 posts

Posted 09 July 2017 - 02:01 PM

Or at what range a laser vomit Warhammer loses to a laser vomit Timberwolf? Or what would happen if 20 meta clan mechs fought 20 meta IS mechs? Do you like to stare at blinking lights and changing numbers?

Enter, the

MWO Loadout Simulator










Posted Image











Add mech loadouts as Smurfy URLs, set the engagement range, and run the simulation. You can change the simulation speed at any time by pressing the 1x to 8x buttons. It's all on the browser, so you don't need to install or download anything.

After one team loses, it'll give a report of how much damage was dealt by each mech and each weapon type. You can also share your simulation by making a permalink other people can view.

It simulates most of the damage and heat mechanics of the game (including ghost heat, loss of heatsinks/ammo/weapons on component destruction, clan XL engine degradation on side torso loss, death on IS XL engine side torso loss) and gives you an idea of how a mech's weapons fire and how much heat builds up when you use it. It also simulates weapon spread on most spread weapons (SRMs, LRMs, Streaks. LBX in progress).

Running a simulation gives you a much better picture of what is more likely to happen in the battlefield than comparing DPS and armor numbers (and is also much cooler looking).

Runs best on Chrome (and other webkit browsers like Opera and Safari). Will run on Firefox, but not well. DO NOT EVEN TRY RUNNING IT ON EDGE OR IE. The simulation is all run on Javascript so your browser will eat a lot of CPU while it's running, especially for large matchups.

You can access it here. Quick guide on how to use the simulation and a summary of the project here. Bug reports are welcome (especially on errors on the implementation of game mechanics, I want to make the simulation as close to the game mechanics as possible).


Here's a bunch of interesting matchups:


Clan vs IS laser vomit
(try it once with the default settings, then change the Clan team's component targeting to 'Aim for XL Side Torso' under settings)
200m 400m 600m


Clan vs IS SRMboats
(again, try once on default, then change targeting to aim for XL side torso)
200m


Warhammer vs Timberwolf DPS
200m 400m 600m


Medium Brawlers DPS
100m 200m


Annoying Lights DPS
100m 200m


Tier 1 Meta DPS Contest
Tier 1 from metamechs.com shooting dummy Atlases
200m 400m 600m 800m


20 vs 20 Clan vs IS Meta Extravaganza
Top IS and Clan mechs from the metamechs tier list. As with all Clan and IS matchups, try it once with default settings, and again with Aim for XL side torso. May run slow on non-Chrome browsers.
200m


And as for the Dire Wolf and Urbanmech question. The answer is four.


Have fun, and feedback is welcome.

Updates:
2017-07-13: Support for Clan and IS weapon quirks (except for Omnipod full set bonuses). The sim will be undergoing active development for the next week, so press Ctrl-F5 (on desktop browsers) to ensure that you always get the latest code. On mobile, you can access the site in incognito mode to make sure it doesn't use old cached versions of the code.
2017-07-17: Implemented UAC double tap. Remember to use Ctrl-F5 to update to the latest code.

Edited by fat4eyes, 17 July 2017 - 12:21 AM.


#2 Y E O N N E

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 16,810 posts

Posted 09 July 2017 - 02:10 PM

Pretty awesome.

Question, though: is this simulator assuming that two competing 'Mechs are just starting straight ahead as they fire into each other, or is there some stochastic estimate on the effects of torso twisting and hit-boxes, etc.?

Edit: I would also suggest a change to the way accuracy is handled. At T1, it says 85% goes to one component while the remaining 15% goes to adjacent. But that's not really what happens. A laser with a duration at or below 0.8 seconds is really hard to spread, regardless of skill, because you are now running into limitations of not just human reaction time but also how fast the 'Mech can physically move. So something like a Large Pulse will not spread the same proportion of its damage as an ER Large Laser or whatever. Does the simulation look at that, too? It seems to me that the most accurate way to test this would be to set a Clan team one tier below the IS team to try and capture the difference in spread/duration between Clan and IS weapons.

Edited by Yeonne Greene, 09 July 2017 - 02:29 PM.


#3 Sadist Cain

    Member

  • PipPipPipPipPipPipPip
  • Philanthropist
  • 605 posts

Posted 09 July 2017 - 02:23 PM

Bravo I say, Bravo.

#4 LordNothing

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 17,175 posts

Posted 09 July 2017 - 02:28 PM

in my experience, one.

#5 r4plez

    Member

  • PipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 812 posts
  • LocationFoundry

Posted 09 July 2017 - 02:28 PM

Works best when in work - thanks OP :D

#6 Tarogato

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Civil Servant
  • Civil Servant
  • 6,558 posts
  • LocationUSA

Posted 09 July 2017 - 02:37 PM

Okay, this is really neat. Definitely looking into this a little deeper to see how useful it really is.


There needs to be two "Max damage per heat" options. One for "chainfire" which is the presently used scheme and not very useful, and one for alpha, where it always waits for the largest non-ghost heat alpha it can fire without overheating, which is how real players behave.


This doesn't factor quirks at all, does it?



View PostYeonne Greene, on 09 July 2017 - 02:10 PM, said:

Question, though: is this simulator assuming that two competing 'Mechs are just starting straight ahead as they fire into each other, or is there some stochastic estimate on the effects of torso twisting and hit-boxes, etc.?


Seems to be standing still shooting each other in the face. But you can set the "accuracy" per team.

#7 Jay Leon Hart

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Spear
  • The Spear
  • 4,669 posts

Posted 09 July 2017 - 02:40 PM

I can not 'like' this enough! Posted Image

#8 Y E O N N E

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 16,810 posts

Posted 09 July 2017 - 02:42 PM

View PostTarogato, on 09 July 2017 - 02:37 PM, said:

Seems to be standing still shooting each other in the face. But you can set the "accuracy" per team.


Already addressed that; the accuracy system is unsatisfactory. You have to set a Clan team to have a worse level of accuracy to account for the longer duration and worse spread, but that doesn't really account for weapons within a tech tree.

#9 Tarogato

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Civil Servant
  • Civil Servant
  • 6,558 posts
  • LocationUSA

Posted 09 July 2017 - 02:44 PM

View PostYeonne Greene, on 09 July 2017 - 02:42 PM, said:


Already addressed that; the accuracy system is unsatisfactory. You have to set a Clan team to have a worse level of accuracy to account for the longer duration and worse spread, but that doesn't really account for weapons within a tech tree.


Well, burn duration and such doesn't matter when mechs are standing still and not torso twisting.


The next step (after adding quirks and skill trees to the simulation... which is a humongous task), would be simulating torso twisting, and then durations would matter. But that also might not even be possible with the code as is, would require a complete rewrite, lol.

Oh, and adding UAC double taps and jamming. That's gotta be high priority.

Edited by Tarogato, 09 July 2017 - 02:53 PM.


#10 Pur

    Member

  • PipPipPipPipPip
  • 188 posts

Posted 09 July 2017 - 02:47 PM

View Postfat4eyes, on 09 July 2017 - 02:01 PM, said:

Medium Brawlers DPS
200m


I clicked the above, hit run, and this is what the outcome was:

Posted Image

#11 Y E O N N E

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 16,810 posts

Posted 09 July 2017 - 02:52 PM

View PostTarogato, on 09 July 2017 - 02:44 PM, said:

Well, burn duration and such doesn't matter when mechs are standing still and not torso twisting.


Spread does. :P

Quote

The next step (after adding quirks and skill trees to the simulation... which is a humongous task), would be simulating torso twisting, and then durations would matter. But that also might not even be possible with the code as is, would require a complete rewrite, lol.


Yeah.

I think it is taking quirks into account, because if you line up one Nova against one Atlas D from the Medium Brawler DPS preset linked above, it takes 217 damage to down that Atlas using perfect accuracy and aiming for CT. Furthermore, the IS seem to usually come out on top when the fight is in their range and CT aiming is set, suggesting more hitpoints overall. Meanwhile, if you set it to target XLs, IS get rekt.

Interestingly (or not, dunno), I ran that 20 v 20 at 200 meters with Clan aiming at T2 and aiming for XL and IS aiming at T1 and CT, and IS still won.

#12 Tarogato

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Civil Servant
  • Civil Servant
  • 6,558 posts
  • LocationUSA

Posted 09 July 2017 - 03:11 PM

View PostYeonne Greene, on 09 July 2017 - 02:52 PM, said:

I think it is taking quirks into account, because if you line up one Nova against one Atlas D from the Medium Brawler DPS preset linked above, it takes 217 damage to down that Atlas using perfect accuracy and aiming for CT. Furthermore, the IS seem to usually come out on top when the fight is in their range and CT aiming is set, suggesting more hitpoints overall. Meanwhile, if you set it to target XLs, IS get rekt.



Ah, interesting. So it actually shows you the health of each mechs components. This bit here:
Posted Image

That could use some labels, indicating first row is armour and second row is structure. And then onhover showing how much it gets from quirks. And it does seem to include structure quirks. But it doesn't include omnipod quirks (the GAR right arms should be very different here, but they aren't)



But I tested a weapon quirk, SDR-5V has 25% energy cooldown, JVN-11A has none. But in the simulator they fire the same speed, no difference:

http://www.4eye-labs...45edf69d9ccfdb0


So not all quirks are working.

Edited by Tarogato, 09 July 2017 - 03:25 PM.


#13 Foxfire kadrpg

    Member

  • PipPipPipPipPipPip
  • The Territorial
  • The Territorial
  • 291 posts

Posted 09 July 2017 - 03:13 PM

I like staring at blinking lights and changing numbers.

#14 Tarogato

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Civil Servant
  • Civil Servant
  • 6,558 posts
  • LocationUSA

Posted 09 July 2017 - 03:15 PM

lollolol, here's the VIP atlas if anybody cares to use it for anything: http://www.4eye-labs...5cfebe1cd68db46

#15 HGAK47

    Member

  • PipPipPipPipPipPipPip
  • Shredder
  • Shredder
  • 971 posts

Posted 09 July 2017 - 03:28 PM

Very interesting, I just spent 10 mins playing around with the IS vs Clan 2 v 2 SRM boat fight. Playing around with the battle ranges n stuff.

I imagine this could provide some neat stats for the people who can utilise this better than me.

InB4 - "Derp this simulator fights better than I do"

Edited by HGAK47, 09 July 2017 - 03:29 PM.


#16 LordNothing

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 17,175 posts

Posted 09 July 2017 - 04:19 PM

proof that cac2 is better than cuac2

at least when you set maximum fire rate and t2 accuracy, apparently the sim settings arent saved in the links.

also the uacs werent jamming so i guess that feature doesnt exist yet. ive tried different sim settings and i either get a draw or a the cac2 wins.

Edited by LordNothing, 09 July 2017 - 04:30 PM.


#17 WrathOfDeadguy

    Member

  • PipPipPipPipPipPipPipPip
  • The Pest
  • The Pest
  • 1,951 posts

Posted 09 July 2017 - 04:39 PM

Ooo, yet another way to play Armchairwarrior during those extended periods of not playing the actual game!

One thing this could be very useful for is testing the relative efficiency of slight build modifications by pitting the updated build against the old one before committing the cbills for new hardware.

Bookmarked. Gonna be spending a bunch of time messing with this.

#18 Y E O N N E

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • The Nimble
  • The Nimble
  • 16,810 posts

Posted 09 July 2017 - 04:59 PM

Would also be interesting to see the impacts certain skill tree allocations has...

#19 jjm1

    Member

  • PipPipPipPipPipPipPipPip
  • Hell Fork
  • Hell Fork
  • 1,384 posts

Posted 09 July 2017 - 05:55 PM

PSA: You can simulate Frozen City by setting the range to 1200

#20 Tarogato

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Civil Servant
  • Civil Servant
  • 6,558 posts
  • LocationUSA

Posted 09 July 2017 - 06:24 PM

View PostLordNothing, on 09 July 2017 - 04:19 PM, said:

proof that cac2 is better than cuac2

at least when you set maximum fire rate and t2 accuracy, apparently the sim settings arent saved in the links.

also the uacs werent jamming so i guess that feature doesnt exist yet. ive tried different sim settings and i either get a draw or a the cac2 wins.


UACs don't double-tap in the simulator yet.



Also, I got this, lol

Posted Image

Edited by Tarogato, 09 July 2017 - 06:25 PM.






17 user(s) are reading this topic

0 members, 17 guests, 0 anonymous users