Jump to content

Worthy Adversaries - What Can Explain This?


112 replies to this topic

#81 RussianWolf

    Member

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

Posted 18 August 2014 - 09:22 AM

View PostRoland, on 18 August 2014 - 09:12 AM, said:

Yeah, you should have opted in, and backed up your talk by showing everyone how secretly awesome the IS heavies are.

But, alas, you didn't. And I think everyone knows why.

All you have to do is press "launch". I'm on almost every night. Played with and against many of the people on the forums here, most will tell you I'm a pretty decent player.

#82 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 18 August 2014 - 09:24 AM

View PostRussianWolf, on 18 August 2014 - 09:22 AM, said:

All you have to do is press "launch". I'm on almost every night. Played with and against many of the people on the forums here, most will tell you I'm a pretty decent player.

Being a decent player doesn't mean you're gonna match clan mech output with an IS mech. That's the point.

But hey, you can still opt in and put yourself up on that leaderboard to show everyone what a real pilot can do. The tournament is still running all day.

#83 PappySmurf

    Member

  • PipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • 842 posts

Posted 18 August 2014 - 09:28 AM

RussianWolf (the data is flawed.

Assumption. All players are equal - then the data would be valid.)

There is only 1 problem with this statement the server AI. Most players do not realize the server AI also controls damage output to achieve a balanced state 1.0.You can see the server AI at work in private duels played multiple times with the same opponent. If 1 opponent kills the other twice his damage output is lowered by the server AI until the killed player kills his opponent twice to achieve the 1.0 balance.

I did not really believe this was true until a group of us tested it for ourselves so with this new understanding of the actual workings of the MWO server AI how can any data gained from PGI staff or Elite players be valid unless the server AI was shut off ?

Plus how are the top 20 leaders on the tournament leaderboard per category IS or Clan gamming the system to achieve there winning result?

http://research.micr...NOSSDAV2007.pdf

Posted Image

Edited by PappySmurf, 18 August 2014 - 09:47 AM.


#84 Clit Beastwood

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,262 posts
  • LocationSouthern California

Posted 18 August 2014 - 09:32 AM

View PostCarrie Harder, on 16 August 2014 - 07:27 PM, said:

Oh, and also, the tourney leaderboard indicates that IS lights are apparently superior to IS heavies. Wut.


the pilots are just better
:ducking:

#85 RussianWolf

    Member

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

Posted 18 August 2014 - 09:45 AM

View PostRoland, on 18 August 2014 - 09:24 AM, said:

Being a decent player doesn't mean you're gonna match clan mech output with an IS mech. That's the point.

But hey, you can still opt in and put yourself up on that leaderboard to show everyone what a real pilot can do. The tournament is still running all day.

Not interested... but if you want to meet up, I'll bring one of my IS builds and you can bring a Twolf. Win or lose, I'll have fun.

#86 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 18 August 2014 - 09:52 AM

View PostPappySmurf, on 18 August 2014 - 09:28 AM, said:

RussianWolf (the data is flawed.

Assumption. All players are equal - then the data would be valid.)

There is only 1 problem with this statement the server AI. Most players do not realize the server AI also controls damage output to achieve a balanced state 1.0.You can see the server AI at work in private duels played multiple times with the same opponent. If 1 opponent kills the other twice his damage output is lowered by the server AI until the killed player kills his opponent twice to achieve the 1.0 balance.

I did not really believe this was true until a group of us tested it for ourselves so with this new understanding of the actual workings of the MWO server AI how can any data gained from PGI staff or Elite players be valid unless the server AI was shut off ?

Plus how are the top 20 leaders on the tournament leaderboard per category IS or Clan gamming the system to achieve there winning result?

http://research.micr...NOSSDAV2007.pdf

Posted Image

You're joking with all this, right?

#87 Carrie Harder

    Clone

  • PipPipPipPipPipPipPip
  • 678 posts
  • LocationCarrying pugs up Mount Tryhard

Posted 18 August 2014 - 10:19 AM

View PostVassago Rain, on 18 August 2014 - 05:46 AM, said:


We do know, because Russ posted all the data. He said elo was very much in range, and the mechs weren't getting mismatched.

Clans are OP P2W. Accept it.

The issue is that PGI probably hasn't yet identified the exact elements of the Clan lineup that caused the 90% skew. Was it the LB 2-X, the SSRM4, Small Pulse Laser, or UAC/10? Was it Adders or Summoners? PGI is going to almost certianly nerf every single aspect of the Clans right down to Pulse Lasers and the "oversized lasers" known as CUACs, even though the issues lie within only 2-3 mechs and like 5-6 weapons.

Edited by Carrie Harder, 18 August 2014 - 10:20 AM.


#88 Malcolm Vordermark

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,520 posts

Posted 18 August 2014 - 10:26 AM

View PostCarrie Harder, on 18 August 2014 - 10:19 AM, said:

The issue is that PGI probably hasn't yet identified the exact elements of the Clan lineup that caused the 90% skew. Was it the LB 2-X, the SSRM4, Small Pulse Laser, or UAC/10? Was it Adders or Summoners? PGI is going to almost certianly nerf every single aspect of the Clans right down to Pulse Lasers and the "oversized lasers" known as CUACs, even though the issues lie within only 2-3 mechs and like 5-6 weapons.


They would likely start with the best performing weapons and mechs. I do not doubt they have the information to discern which ones they are. Also, I doubt they will nerf pulse lasers. I think they will nerf ER lasers until you have to take pulse. That would help a lot with the range advantage and tie up some of that weight savings from the XL. Not saying I like that way but after the ER LL nerf I would not be surprised.

#89 Carrie Harder

    Clone

  • PipPipPipPipPipPipPip
  • 678 posts
  • LocationCarrying pugs up Mount Tryhard

Posted 18 August 2014 - 10:28 AM

View PostRouken, on 18 August 2014 - 10:26 AM, said:


They would likely start with the best performing weapons and mechs. I do not doubt they have the information to discern which ones they are. Also, I doubt they will nerf pulse lasers. I think they will nerf ER lasers until you have to take pulse. That would help a lot with the range advantage and tie up some of that weight savings from the XL. Not saying I like that way but after the ER LL nerf I would not be surprised.

And then when people start taking Pulse due to overnerfed ER lasers, then they'll nerf the Pulses. :) It's been shown that they don't always interpret data effectively.

It's the circle of meta. Nerf one thing, another takes it place, then nerf the new thing, etc....

#90 LuInRei

    Member

  • PipPipPipPipPip
  • Philanthropist
  • 167 posts
  • Location渦巻き中

Posted 18 August 2014 - 10:36 AM

View PostCarrie Harder, on 18 August 2014 - 10:28 AM, said:

And then when people start taking Pulse due to overnerfed ER lasers, then they'll nerf the Pulses. :) It's been shown that they don't always interpret data effectively.

It's the circle of meta. Nerf one thing, another takes it place, then nerf the new thing, etc....



It is the direct result and consequence of balancing by server statisctics,
which, I fear, is the approach employed by PGI regarding balance.
Whatever weapon or mech shows signs of increasing performance is nerfed,
and things that are not used or showing bad stats are buffed.

#91 Clit Beastwood

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,262 posts
  • LocationSouthern California

Posted 18 August 2014 - 10:39 AM

View PostPappySmurf, on 18 August 2014 - 09:28 AM, said:

RussianWolf (the data is flawed.

Assumption. All players are equal - then the data would be valid.)

There is only 1 problem with this statement the server AI. Most players do not realize the server AI also controls damage output to achieve a balanced state 1.0.You can see the server AI at work in private duels played multiple times with the same opponent. If 1 opponent kills the other twice his damage output is lowered by the server AI until the killed player kills his opponent twice to achieve the 1.0 balance.

I did not really believe this was true until a group of us tested it for ourselves so with this new understanding of the actual workings of the MWO server AI how can any data gained from PGI staff or Elite players be valid unless the server AI was shut off ?

Plus how are the top 20 leaders on the tournament leaderboard per category IS or Clan gamming the system to achieve there winning result?

http://research.micr...NOSSDAV2007.pdf

Posted Image



? not gaming the system, did 20 drops in lights and wound up on the boards :shrug: - I do feel like the lasers on my IS lights do way less damage than they used to, but :shrug: whatever :)

#92 Lefty Lucy

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 3,924 posts
  • LocationFree Tikonov Republic

Posted 18 August 2014 - 11:05 AM

View PostFierostetz, on 18 August 2014 - 10:39 AM, said:



? not gaming the system, did 20 drops in lights and wound up on the boards :shrug: - I do feel like the lasers on my IS lights do way less damage than they used to, but :shrug: whatever :)


Yeah, hitreg against moving targets right now is a mess. I'll get a full second burn on a light's red-armor leg without getting to internals with 4 ML With the same shot, however, I'll get right through the rear armor on a DWF.

#93 Aresye

    Member

  • PipPipPipPipPipPipPipPipPip
  • Heavy Lifter
  • Heavy Lifter
  • 3,462 posts

Posted 18 August 2014 - 11:08 AM

View PostVassago Rain, on 18 August 2014 - 05:46 AM, said:


We do know, because Russ posted all the data. He said elo was very much in range, and the mechs weren't getting mismatched.

Clans are OP P2W. Accept it.


To be quite honest, it's getting old really quick seeing the same people like you and Roland taking every opportunity to derail any thread that merely mentions clan mechs into the same argument.

You're not even arguing against people who are vehemently defending Clan mechs like in the past. You're arguing against those (including myself) who agree that there are certain chassis and weapons that need to be toned down. The only logical reason that one would argue against those trying to come up with suggestions to actually balance the two out would be that they themselves are not concerned with balance, and Clans will never be considered balanced in their eyes unless IS mechs are vastly superior to them.

Russ threw your crowd a little treat with the 90% statistic, and while I "agree" that this does indicate the Clans do have an advantage and will require some tweaking, it has absolutely NO indication of just how superior Clan mechs are compared to IS. All the 90% statistic says is, "Clans have an advantage." Nothing more. Nothing less. For all we know the advantage could be relatively small, but just enough to tip the scales to produce a skewed result.

Take a car race for example. One car can go 5mph faster than the other cars. It may win 90% of the races, but its overall advantage is relatively small. You're assuming that because that one car wins 90% of the races, that it must be going at least 40mph faster than the other cars, and the only way to balance it would be to make it go 45mph slower, which would end up making it so that car can no longer be able to compete in the first place.

The closest thing that we have to hard data on actual performance between Clan and IS was compiled by Kiiyor, which puts the Clans at a 16% advantage in overall damage, and a 14% advantage in overall kills: http://mwomercs.com/...ew-data-170714/

I get it though. Saying Clans have a 15% advantage over IS sure doesn't sound as horrible and shocking as claiming that they have a 90% advantage.

#94 Kassatsu

    Member

  • PipPipPipPipPipPipPipPip
  • The Devoted
  • The Devoted
  • 1,078 posts
  • LocationColorado

Posted 18 August 2014 - 11:17 AM

View PostJeb, on 17 August 2014 - 12:20 AM, said:

otherwise your gonna get a bunch of crappy bandaids tossed in that just break more things..


Too bad that's been the stance up until now. I don't expect anything less than an all-around 33% nerf to random things people weren't even whining making balance suggestions about.

Are clans OP? Yes. Do they need their weapons curbstomped so hard they can't even compete with their IS counterparts? No.

#95 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 18 August 2014 - 11:54 AM

View PostRhialto, on 16 August 2014 - 07:18 PM, said:

I look at the Heavy leaderboard and notice that the 15th from Clan have more points than the very 1st from IS. Also, 2,711 for the 1st in Clan is a lot more higher than the 1st IS with 2,115.

What can explain this?

Clan Omnis have been delivering more damage that their IS counter parts for going on 30 years. It is nothing new.

#96 Johnny Z

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Knight Errant
  • 9,942 posts
  • LocationDueling on Solaris

Posted 18 August 2014 - 01:35 PM

View PostKassatsu, on 18 August 2014 - 11:17 AM, said:


Too bad that's been the stance up until now. I don't expect anything less than an all-around 33% nerf to random things people weren't even whining making balance suggestions about.

Are clans OP? Yes. Do they need their weapons curbstomped so hard they can't even compete with their IS counterparts? No.


I'm not so sure about that, maybe a week or 2 of IS OP would change the tunes of the Clans that think OP is just fine. :)

#97 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 18 August 2014 - 01:42 PM

View PostKassatsu, on 18 August 2014 - 11:17 AM, said:


Too bad that's been the stance up until now. I don't expect anything less than an all-around 33% nerf to random things people weren't even whining making balance suggestions about.

Are clans OP? Yes. Do they need their weapons curbstomped so hard they can't even compete with their IS counterparts? No.
It depends on what standard you wish to apply Kas. To me Omnis are NOT OP. I don't find them powerful enough to be the boogiemen I have fought and played for the last 20 years.

#98 Johnny Z

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Knight Errant
  • 9,942 posts
  • LocationDueling on Solaris

Posted 18 August 2014 - 01:49 PM

View PostJoseph Mallan, on 18 August 2014 - 01:42 PM, said:

It depends on what standard you wish to apply Kas. To me Omnis are NOT OP. I don't find them powerful enough to be the boogiemen I have fought and played for the last 20 years.


You heard the man game makers, give IS a large larser that shoots further and does more damage than the clans have and one of each of the other weapons as well to be fair. Just for a week or 2 :) Also let the Inner Sphere XL engines lose a side to no effect as well, to be fair. I could go on of course but I cant type because I am giggling to much at the thought. ;)

#99 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 18 August 2014 - 01:55 PM

View PostJohnny Z, on 18 August 2014 - 01:49 PM, said:


You heard the man game makers, give IS a large larser that shoots further and does more damage than the clans have and one of each of the other weapons as well to be fair. Just for a week or 2 :) Also let the Inner Sphere XL engines lose a side to no effect as well, to be fair. I could go on of course but I cant type because I am giggling to much at the thought. ;)

You must have read my post wrong Johnny. I want The Clans more powerful not my Inner Sphere swag!
:unsure:
Or are you being Sarcastic. It's hard to tell which voice to read a post with!

Edited by Joseph Mallan, 18 August 2014 - 01:56 PM.


#100 Vassago Rain

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Bridesmaid
  • Bridesmaid
  • 14,396 posts
  • LocationExodus fleet, HMS Kong Circumflex accent

Posted 18 August 2014 - 02:42 PM

View PostRussianWolf, on 18 August 2014 - 08:15 AM, said:

Well, I didn't opt in but maybe I should have. Played nothing but IS all weekend (and will be for a while yet as I'm grinding some elites) and had no issue what-so-ever in killing Timber Wolves and Dire Wolves using my Inferior mechs.

Did I say the numbers are wrong? No. I said its flawed data so you can't trust it to be reliable. To get reliable data you have to limit or remove as many variables as possible so that you can see the correlation of the one variable you want to check.

Here's how. Take 24 players and put them on 2 teams. Similarly skilled helps, but isn't required. Those two teams are now fixed for the test. Now take 24 mechs (12 clan and 12 IS ) of similar builds (so all should be running DHS, and have similar weapon load out as much as possible. At the end of each match, you swap rolls but stay in the matched mech. And do this as many times as you can comfortably do it before any variable change. So in the end you get to see how the IS v Clan does and you have effectively eliminated or minimized the variable of player and mech type/load out.

Bad input data is where you simple take 12v12, clanvIS and through out the fact that the teams are 1)changing with each match, 2)same players are playing the same side, 3)same players are playing different builds and different classes,

Hard to get reliable data in the end that way.

Their numbers may in fact be correct. But you can't determine that from the way that they are gathering their data.


That's called a biased test.

What they normally do are all biased tests. This last one, for some reason, was a proper test, and Russ - probably by accident - posted about it on twitter. Clans have a 90% win ratio.

Regardless of skill level.

Regardless of ghetto or mountain top.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users