Jump to content

Is Vs Clans, With Science! New Data - 17/07/14

Balance General

312 replies to this topic

#81 DONTOR

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 3,806 posts
  • LocationStuck on a piece of Commando in my Ice Ferret

Posted 27 June 2014 - 09:16 AM

Damn dude I dont think I've ever seen that much time and effort put into a thread before! You went above and beyond, that is some pretty cool info. Having piloted DWs quite a bit since the release the lower kills per game stat is becuase of how slow they are to get to the fight or into a good position to take advantage of a weak mech (more ppl need to realise they are better at range than they are at brawling (CT)).

#82 TexAce

    Member

  • PipPipPipPipPipPipPipPipPip
  • The 1 Percent
  • The 1 Percent
  • 2,861 posts
  • LocationGermany

Posted 27 June 2014 - 09:38 AM

I still wish we would fight 10 clans vs 12 IS and not mixed....

PS: first time I've read through such a long opening post on this forum, speaks about the quality of the OP (and the quality of the others)

Edited by TexAss, 27 June 2014 - 09:40 AM.


#83 101011

    Member

  • PipPipPipPipPipPipPipPip
  • The Infernal
  • The Infernal
  • 1,393 posts
  • LocationSector ZZ9 Plural Z Alpha, on a small blue-green planet orbiting a small, unregarded yellow sun.

Posted 27 June 2014 - 10:09 AM

View PostKiiyor, on 24 June 2014 - 07:23 PM, said:

I wish. I should call for volunteers

Well, I doubt it would be difficult to find guys interested in doing some Clan v. IS matches.

#84 Noober

    Member

  • PipPip
  • Ace Of Spades
  • Ace Of Spades
  • 43 posts

Posted 27 June 2014 - 10:21 AM

Excellent work!!! Thanks for doing this, I've been meaning to do something like this for a while.

Somebody get this man some premium time!

#85 RussianWolf

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 2,097 posts
  • LocationWV

Posted 27 June 2014 - 10:57 AM

View Postravineh, on 27 June 2014 - 09:12 AM, said:

So the average dmg/kills/match is higher for clans. In short: they are better atm.
Pretty close to my ongoing tests. Was planning to make similar statistics once i hit 300 games (IS mech). not far from it, so gonna post later.

or maybe gonna be lazy and mail ss to you, heh


Can't really say Clan mechs are better by that statistic by itself. What are the average numbers for those same players with IS mechs of similar loadout/weight? Picking the one statistic and coming to a conclusion isn't very good science.Too many other variables to take into account.

#86 ravineh

    Member

  • Pip
  • 16 posts

Posted 27 June 2014 - 11:34 AM

View PostRussianWolf, on 27 June 2014 - 10:57 AM, said:


Can't really say Clan mechs are better by that statistic by itself. What are the average numbers for those same players with IS mechs of similar loadout/weight? Picking the one statistic and coming to a conclusion isn't very good science.Too many other variables to take into account.


it's quite easy actually. In theory players in the same match have the same elo (+-) so if they score better (make more dmg/kills) in clan mechs, then yes they perform better

#87 Wispsy

    Member

  • PipPipPipPipPipPipPipPipPip
  • Talon
  • Talon
  • 2,007 posts

Posted 27 June 2014 - 11:36 AM

Clan lights score higher then Jenners...lol?

Which in turn..better then Cataphract...

Edited by Wispsy, 27 June 2014 - 11:39 AM.


#88 RussianWolf

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 2,097 posts
  • LocationWV

Posted 27 June 2014 - 11:44 AM

View Postravineh, on 27 June 2014 - 11:34 AM, said:


it's quite easy actually. In theory players in the same match have the same elo (+-) so if they score better (make more dmg/kills) in clan mechs, then yes they perform better

do you know what the elo spread they are currently using is?

A Catapult Missile Boat A1 can not be compared to a Laser/PPC build TW with any good data being extracted.

So no, its not that simple.

#89 Lynx7725

    Member

  • PipPipPipPipPipPipPipPip
  • 1,710 posts

Posted 27 June 2014 - 11:45 AM

View Postravineh, on 27 June 2014 - 11:34 AM, said:

it's quite easy actually. In theory players in the same match have the same elo (+-) so if they score better (make more dmg/kills) in clan mechs, then yes they perform better

Nope. The teams have the same elo (+-). The players may not.

#90 Why Run

    Member

  • PipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 370 posts

Posted 27 June 2014 - 11:48 AM

Use of average is wrong. Please come back with normalized numbers and medians. We don't want to know average, which we all know is easily skewed. Without standard deviations and some sort of analysis there isn't much you can draw from this...

#91 RussianWolf

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 2,097 posts
  • LocationWV

Posted 27 June 2014 - 11:50 AM

in order to get the best data, you have to either eliminate as many variables as possible or manipulate the variable to a very narrow spectrum.

That why I said you need to be able to follow specific players (one variable down) in both IS and Clan mechs of similar loadout/weight (another variable down). Then you are left with the only variable being Clan vs IS and your data will be better (still not 100% accurate, but at least better).

#92 arcangelS7

    Member

  • PipPip
  • Knight Errant
  • 33 posts

Posted 27 June 2014 - 12:06 PM

Well done OP.
This is the kind of analysis that game balancing should be based on.
I think we would all feel better if PGI release a similar summary of their own internal statsics on a regular basic (weekly prehaps?) to give the community a wider view of the current game situation.

#93 HBizzle

    Member

  • PipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 522 posts
  • LocationDC

Posted 27 June 2014 - 01:12 PM

Thank you for putting all this data together.
Is there a way to send this guy MC?

Key stat in it is damage per side compared to tonnage per side. Right now the IS mechs are 44% of the tonnage per drop, but only doing 39% of the damage. So they are under performing by only 4% compared to their tonnage per drop. That to me shows that the claims Clan mechs are OP is ******* rubbish. Under performing by 4% is barely statistically significant, and with such a small sample size, 100 matches, can even be considered statistical noise. You can argue OMG need median and standard deviation, but ton for ton the IS mechs are not under performing much at all compared to the Clan mechs by a significant amount.


Looks pretty balanced to me, and more 4P action. Piss Poor Pilots Pouting.
Also Comic Sans for flame retardant purposes.

Edited by HBizzle, 27 June 2014 - 01:15 PM.


#94 Khanahar

    Member

  • PipPipPipPipPipPipPip
  • The Bold
  • The Bold
  • 560 posts

Posted 27 June 2014 - 01:34 PM

This is excellent data but I'd love to do more with the numbers (see what they look like with you excluded, check medians, try excluding outliers, see standard deviations, etc.). Is there any chance you could put up the raw data in a google doc or fileshare?

#95 BladeSplint

    Member

  • PipPipPipPipPipPip
  • 290 posts

Posted 27 June 2014 - 01:46 PM

I'll upload some EOG screenshots when I get back from my weekend trip, I have tons.

#96 Angel of Annihilation

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Infernal
  • The Infernal
  • 8,881 posts

Posted 27 June 2014 - 01:49 PM

Your evaluation seems right on in virtually all aspects.

Some things to point out.

You get some average damage spikes in the Clan Medium and Heavy range. This is because the Timberwolf and Stormcrow are probably the two most well designed mechs in the game. They are both very versatile designs that can pretty much be customized to any persons play style and having a mech that just "fits" the way you prefer to play is always going to help with how well your going to play that mech so I am not surprised to see higher than average damage numbers from these mechs.

I also was happy to see where the Quickdraw placed because despite the naysayers, that is one beastly little mech in the right hands. Small sample or not, I absolutely believe the numbers are accurate representations.

One thing I would like to kind of refute is your evaluation of lights not doing so well. First you think have to understand is that a light mech isn't designed to be a damage powerhouse or configured as a mech killer. Lights rather are designed for scouting, recon, harassment and electronic warfare and they excel at this. For example give a couple lights ECM and if (big if here) they are actually covering the main group, they have a huge impact on the outcome of a battle and they don't even have to fire a shot. Another example is how a few lights roaming the backfield can totally stop or disrupt any coordinated enemy advance again without having to do more than fire a few token shots to get people's attention.

Basically lights are supposed to be there for the team's sake so their only problem exists in the minds of light pilots who think that their 5 tons of weapons space should allow them to regularly kill 5 mechs and do 800 damage a match rather than provide ECM coverage for the group or harass the enemy rear to insure the team wins.

Finally as to the main topic, clan vs IS balance, I think we see that overall, Clan mechs, aside from perhaps a few individual builds, really aren't all that superior to their IS counterparts,

#97 Jman5

    Member

  • PipPipPipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 4,914 posts

Posted 27 June 2014 - 01:52 PM

View PostHBizzle, on 27 June 2014 - 01:12 PM, said:

Thank you for putting all this data together.
Is there a way to send this guy MC?

Key stat in it is damage per side compared to tonnage per side. Right now the IS mechs are 44% of the tonnage per drop, but only doing 39% of the damage. So they are under performing by only 4% compared to their tonnage per drop. That to me shows that the claims Clan mechs are OP is ******* rubbish. Under performing by 4% is barely statistically significant, and with such a small sample size, 100 matches, can even be considered statistical noise. You can argue OMG need median and standard deviation, but ton for ton the IS mechs are not under performing much at all compared to the Clan mechs by a significant amount.


Looks pretty balanced to me, and more 4P action. Piss Poor Pilots Pouting.
Also Comic Sans for flame retardant purposes.

I'll repost my reddit response here as well



I'm not sure how you arrived at 4%, but if you do a damage/ton for each side you get
IS: 4.00 D/T
Clan: 4.7877 D/T
That's about an 18% damage differential per ton.

#98 jaxjace

    Member

  • PipPipPipPipPipPipPip
  • The Merciless
  • The Merciless
  • 987 posts
  • LocationIn orbit around your world

Posted 27 June 2014 - 01:54 PM

Stalker will save the day. Again.

#99 HBizzle

    Member

  • PipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 522 posts
  • LocationDC

Posted 27 June 2014 - 01:57 PM

View PostJman5, on 27 June 2014 - 01:52 PM, said:

I'll repost my reddit response here as well



I'm not sure how you arrived at 4%, but if you do a damage/ton for each side you get
IS: 4.00 D/T
Clan: 4.7877 D/T
That's about an 18% damage differential per ton.


And I will repost my response so folks here can see it.

I don't think you got what I was saying, or I wasn't clear enough, probably the later. Two key stats are below.

Total Damage per Faction
Faction Total %
IS 306883 39.91%
Clans 462113 60.09%

Tonnage per Faction
Faction Total %
IS 76695 44.28%
Clans 96520 55.72%

Right now the total Tonnage of all IS vs Clans is a 44.28% to 55.72%. That means if everything was equal, as in damage, the IS and Clans break down of total damage should be 44.28% to 55.72% because that is the split in weight, which is the only metric that is equal between the two. Instead the percentage of damage is 39.91% to 60.09%, which means the difference between the weight and expected damage is -4.37% against the IS.

That is the key stat as far as balance goes, because if things were perfectly balanced, which they aren't, then the split would be 0% between the difference between damage share and tonnage share. So all we really know right now is that Clan mechs are outperforming their weight share compared to expected damage by only 4.37%, which with such a small sample size isn't statistically significant.

Damage per ton is assuming each ton is the same, which we know it isn't. Damage compared to weight share, or %, is a better estimate since that should be an indicator of expected damage share if both sides were equal.

#100 Peenutts

    Member

  • PipPipPip
  • The Undertaker
  • The Undertaker
  • 69 posts

Posted 27 June 2014 - 02:20 PM

You should get a free mech bay for the sheer amount of time you invested in doing what ever it is this is supposed to do.
Bravo! :P





13 user(s) are reading this topic

0 members, 13 guests, 0 anonymous users