Jump to content

Clan Vs Is Match Maker Testing Today


211 replies to this topic

#181 Kraven Kor

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 5,434 posts

Posted 19 September 2014 - 05:58 AM

View PostInvisibleTank, on 18 September 2014 - 04:31 PM, said:

I'll throw in my vote for 10 vs 12. If anything, just try it and see how it goes. Get some data and reactions from people. Hopefully PGI has kept their new MM code decoupled enough to allow this with a trivial configuration change. My guess is there's something in the legacy code that is making it more difficult...


I personally think they need to rethink Matchmaker and ELO to make 10 v 12 a possibility.

But I'm not insistent on that. Maybe another workable solution is ensuring that IS has a tonnage advantage.

Or maybe in CW IS will get more reinforcements than clan, and / or get reinforcements sooner.

As a devoted IS player; I sincerely do want to be fighting the clans at a tech disadvantage. I do not want to see clan tech watered down such that we are "equal but different." I do like the idea of IS mechs having quirks to offset that, and I'm not going to get mad if they can't do 10 v 12, but it would be my preference to stick to the lore as much as possible.

Another thought is in CW making IS almost always be on the defense, is giving IS side defense turrets and - maybe, someday - consumables that summon non-mech assets; an Inferno SRM squad or some light tanks, more an annoyance than a real threat but used smartly can help turn the tide.

And yes, if IS can "summon" some infantry or combined arms, Clans should be able to summon Elementals.

Note: For Elementals to be added, I must be able to grab said Elementals and throw them at Clan mechs as improvised weapons.

#182 Procyon Alpha

    Member

  • PipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 165 posts
  • LocationIndianapolis, IN USA

Posted 19 September 2014 - 06:12 AM

10 V 12 is not going to happen at this point so we need to move on and look at other options. 10 V 12 is never going to work in the pub que anyway. I would like to see ELO redone into something like XVM so we know what kind of score we do have and a way to calculate them.

#183 Hoax415

    Member

  • PipPipPipPipPipPipPip
  • 645 posts

Posted 19 September 2014 - 06:59 AM

Agreed.

10v12 is completely dead as an idea and there are legit reasons for that and it is time to move on.

So we switch over to, can we use tonnage instead of weight classes to balance teams or is that impossible too?

If we aren't getting any new MM or team restriction systems in time for CW because of reasons then the obvious method to get that last bit of balance for Clan vs IS in CW look something like this:

IS = 2L/4M/3H/3A versus Clan = 3L/4M/4H/1A

You would expect a setup like this would help the IS winrate a little, no nerfs required.

Edited by Hoax415, 19 September 2014 - 07:00 AM.


#184 GenJack

    Member

  • PipPipPipPipPipPip
  • The God
  • The God
  • 271 posts

Posted 19 September 2014 - 07:00 AM

View PostRuss Bullock, on 16 September 2014 - 04:40 PM, said:

Okay I think I am going to bring the test down early at 6pm, we did manage to get some good test data - basically enough.

A few were upset about the test I guess railroading their night with matches that are unfair. A misjudgetment on my part I suppose, I thought everyone would love it. Perhaps most did.

Anyhow going to edit the original post and have the guys bring it down in 20 minutes.



I really enjoyed this actually!

It was fun, the good hard fought matches that I started to see was when the IS started to kind of play like the IS and the Clanners seemed to enjoy their speed over tactics and spread out too far in a lot of cases which allowed IS Lances to kill one at a time.

That being said ... I still think it should be 10 v 12 ... as at the end of the day I had seen more clan wins then IS.

#185 Kraven Kor

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 5,434 posts

Posted 19 September 2014 - 12:22 PM

And, while it may not even be possible, I'd love to see clanners somehow held to clan honor system.

Some mechanic where they select or are assigned a target and are rewarded for single-handedly defeating that target while they lose rewards if they assist other clan mechs.

Again - may not even be possible, but something I'd like to see.

#186 carl kerensky

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 395 posts
  • LocationMoon Base Alpha

Posted 19 September 2014 - 01:47 PM

View PostRuss Bullock, on 16 September 2014 - 01:47 PM, said:

I have decided to turn the Match Maker to Clan vs IS matching for a few hours this afternoon/evening. I think it is time that we collect some new data on the current state of balance between the two technologies.

We will of course be keeping player skill and other factors in mind when looking at the data, while compariing it to the previous IS vs Clan tests.

When: 3:00 PM PDT and will last for an estimated 3 hours depending on results and the quality of play.

Expected Behaviour: Players should expect to experience Clan vs IS matches or when necessary the MM will release valve IS vs IS or Clan vs Clan matches.

I may also consider creating a test window tomorrow during our peak European times.

Update: Solo Queue only

Update 2: changed the duration to 3 hours to restore more equality of play.


Ya good call Russ. Test it please. I like the idea of Lore driven clan vs. IS.

edit whoa i was late for the party lol

Edited by carl kerensky, 21 September 2014 - 06:57 AM.


#187 John McOrigin

    Member

  • Pip
  • 11 posts

Posted 19 September 2014 - 04:33 PM

I think the Matchmaker needs to consider the elo even more: just group up players of a similar elo. The differences in player skills in the same team are too big.

#188 DeathlyEyes

    Member

  • PipPipPipPipPipPipPip
  • The Messenger
  • 940 posts
  • Google+: Link
  • Facebook: Link
  • Twitter: Link
  • Twitch: Link
  • LocationMetaphorical Island somewhere in the Pacific

Posted 20 September 2014 - 03:30 AM

clan vs IS is fairly balanced. The problem is all the trial IS mechs running around. huge case of L2P

PPCs are another issue. Set them back and the IS will once again have an advantage. IS PPCs were one of their best weapons now they are next to useless.

Edited by DeathlyEyes, 20 September 2014 - 03:31 AM.


#189 Kirkland Langue

    Member

  • PipPipPipPipPipPipPipPip
  • Bad Company
  • 1,581 posts

Posted 20 September 2014 - 06:09 AM

And I'll say again, Russ should turn the MM back and run it like that until CW comes out.

TBH, I expect Russ and company are off wasting their time on their MMO more than worried about balance and changes to MWO. So many things should and could be changed in MWO instantly - but nothing happens. You want to balance IS and Clan? How about just turn on the MM and then daily tweak things until it feels right? Instead PGI will wait a few months then make a big deal about rolling out "Clan/IS" balances which in reality are nothing more than tweaks that took all of an hour to implement.

#190 JohnnyWayne

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,629 posts

Posted 20 September 2014 - 07:26 AM

View PostDeathlyEyes, on 20 September 2014 - 03:30 AM, said:

clan vs IS is fairly balanced. The problem is all the trial IS mechs running around. huge case of L2P

PPCs are another issue. Set them back and the IS will once again have an advantage. IS PPCs were one of their best weapons now they are next to useless.


To be honest... I'd love if modules would do actually something intresting. This is a good example. Input a module based trigger to the former made changes to PPC / Gauss desynch. Give the player a module that buffs the speed of PPCs to 2000 and forbid them to fire Gauss the next 4 seconds.

People that would want and could handle the increased complexity could use this. Current and former Pinpoint issues balanced, gg.

Edited by JohnnyWayne, 20 September 2014 - 07:29 AM.


#191 Inti Raymi

    Member

  • PipPipPip
  • 75 posts

Posted 20 September 2014 - 08:25 AM

View PostDeathlyEyes, on 20 September 2014 - 03:30 AM, said:

clan vs IS is fairly balanced. The problem is all the trial IS mechs running around. huge case of L2P

PPCs are another issue. Set them back and the IS will once again have an advantage. IS PPCs were one of their best weapons now they are next to useless.


Facts be damned, I'm right.

There are now clan trial mechs. Noobs (like me) pick them as well. In fact, my guess is that every noob that had the chance to take one did. Since the balance still tilts towards Clan, things are obviously well balanced. If you bought the clan mech packs, that is.

#192 RAM

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Resolute
  • The Resolute
  • 2,019 posts
  • Google+: Link
  • LocationVancouver, BC

Posted 20 September 2014 - 12:38 PM

View PostRuss Bullock, on 16 September 2014 - 04:48 PM, said:

Here is the basic data collected from a sample of 268 IS vs Clan matches.

Clan win rate of 73%

This is a reduction from 90%


Except again it is a meaningless statistic from a pointless test. Your methodology is fundamentally flawed and therefore CANNOT provide ANY useful data, let alone anything sufficiently reliable to perform analysis on.

Not enough matches
No control
No isolation of variables (maps, mechs, players)
Trial vs Customized


RAM
ELH

#193 Hoax415

    Member

  • PipPipPipPipPipPipPip
  • 645 posts

Posted 20 September 2014 - 01:43 PM

That is not how data collection works. At all.

All data collection that isn't purely from a double blind trial with every variable thought of and controlled is not worthless.

Far from it.

If I was hearing bullshit artists like you push for a full challenge weekend with the group queue turned to IS vs Clan... I would respect that and join the chorus.

But I'm not. You want to ignore the data. You want to ignore the competitive players. You want to pretend we can't just look at smurfy and see how you get more weapons on a faster clan mech if it has endo+ff than you do on a same ton IS mech.

And because you say so its balanced.

One side of this "debate" has data, which is far from perfect but its something. The other side has nothing but wishes and denial of reality.

Edited by Hoax415, 20 September 2014 - 01:44 PM.


#194 Jeb

    Member

  • PipPipPipPipPipPip
  • Philanthropist
  • Philanthropist
  • 441 posts
  • Twitter: Link
  • LocationHalifax

Posted 20 September 2014 - 05:29 PM

View PostWintersdark, on 16 September 2014 - 06:46 PM, said:

I saw one fellow, who shall remain nameless due to Name And Shame policies, constantly crying "We've lost, die fast", "it's over, get out of the match" and other such things - then killing himself/getting himself killed quickly in every match we faced Clan mechs.

Several others repeatedly doing much the same.




With that said, I'm absolutely up for IS mechs getting some buffy quirks. They'd really help spice things up! So, I'm quite happy with Russ' response.

I was just about to post almost the same thing...
I only got into a few clan vs IS matches, and I was in a kitfox without ecm...
Few problems:
-one match had 4 kitfoxes, 3 with ecm.... the IS team had 1 ecm which died early...
-another match had 4 timberwolves, and 2 stormcrows... probably the best two clan mechs currently in the game... I will be curious to see if we see more variety once the new clan mechs come out which might help some balance too...

Now here is the crazy thing... in one match, right at the start pretty much, an IS heavy comes running into the middle of the clan mechs and just stopped and stood there... I have seen lights run into a pack of enemys and try to distract them, but this guy didn't even try to fight. That meant the IS team was down a heavy from the start.

And on another match, a heavy and a medium did almost the same thing.. .clan got 3 kills early on (2 lights and a medium that tried to flank or something) and we had then moved on towards their main force... we come around the corner and the medium and heavy started to attack, then just stopped and stared at us... they were not disconnected as far as I could tell as they were moving a bit, but they almost just looked like they gave up and waited to die...

If I saw that in 2 of the 4 or 5 matches I played, then how many other matches had the same thing happen, which would mess with the win/loss stats?

#195 Sabazial

    Member

  • PipPipPipPipPipPipPip
  • Point Commander
  • Point Commander
  • 725 posts
  • LocationUnited Kingdom

Posted 20 September 2014 - 05:31 PM

Personally, i think this "test" is skewed unfavourably toward nerfing clan players. Simply because of one reason, most new players in trial mechs or new starters just getting into the game tend to pick inner sphere mechs first. This is causing a higher number of learning / not as skilled players in the IS queue. Most vets are playing clan mechs currently (other than those die hard IS junkies) so it stands to reason that these IS teams are going to get rolled consistently by said vets no matter what type of mech they choose to play.

I personally played during this test and saw multiple IS mechs during the course of the night with issues such as wandering around like they had no clue of the map layout, guys trying to shoot enemy mechs from 1000m + out with machine guns until they ran out of ammo, players trying to use SRMs like LRMs, people trying to play using gamepads etc, the list can go on but you get the picture. I have also seen the same thing more frequently since around a week and a half before this latest test and it's continuing as there seems to be an influx of newbies recently.

How can this test honestly be impartial if these guys are getting stomped by vets in clan mechs? When the clans went up against IS players of similar skill to their own, they often got their ass handed to them. These tests just feel awfully weighted to me.

#196 Mystere

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 22,783 posts
  • LocationClassified

Posted 20 September 2014 - 06:43 PM

View PostHoax415, on 20 September 2014 - 01:43 PM, said:

That is not how data collection works. At all.

All data collection that isn't purely from a double blind trial with every variable thought of and controlled is not worthless.

Far from it.

If I was hearing bullshit artists like you push for a full challenge weekend with the group queue turned to IS vs Clan... I would respect that and join the chorus.

But I'm not. You want to ignore the data. You want to ignore the competitive players. You want to pretend we can't just look at smurfy and see how you get more weapons on a faster clan mech if it has endo+ff than you do on a same ton IS mech.

And because you say so its balanced.

One side of this "debate" has data, which is far from perfect but its something. The other side has nothing but wishes and denial of reality.


The problem is that the validity of the data is being put into question by the methodology PGI chose. PGI just has to explain why they are conducting the test in that manner.

PGI's methodology looks similar to using a student pilot, a crop duster pilot, an airline pilot, and an air force fighter pilot to determine which is better, a French Rafale or a Russian SU-27 . Do you think that is a valid test? If so, why?

Edited by Mystere, 20 September 2014 - 06:45 PM.


#197 Hoax415

    Member

  • PipPipPipPipPipPipPip
  • 645 posts

Posted 20 September 2014 - 07:05 PM

You already got that entire argument destroyed when several "fighter pilots" came and told you in another thread that there is zero debate in the competitive ranks that clan mechs are stronger.

That argument is further garbage because we're using the MM system to make these matches. There is no reason to think that the elo variance is outside of the normal levels we see in solo and group queue currently. Yes clan teams overall had a 100 elo higher avg. That does not invalidate the testing at all if PGI knows that +100 elo should give (making up numbers) a 55% winrate and instead it gave a 75% winrate. That's statistically significant.

Again a testing method where you know there will be noise in the signal just means you need a larger amount of data to overcome that noise.

So if you were arguing that the 73% test was not very meaningful because they didn't run it long enough or didn't collect enough matches. I think I'd agree, it certainly felt like a much smaller data set then the test that gave us 90%. We should have at least collected as many matches as that time.

In fact if I could make a suggestion to PGI, they should showcase these CW primetime windows they have teased info about by running some Clan vs IS tests during the times when CW will be run.

Edited by Hoax415, 20 September 2014 - 07:07 PM.


#198 Lambulance

    Member

  • PipPip
  • Shredder
  • Shredder
  • 22 posts

Posted 20 September 2014 - 11:09 PM

Give IS melee weapons. Like an energy axe, or maybe even a lightsaber.

#199 Lambulance

    Member

  • PipPip
  • Shredder
  • Shredder
  • 22 posts

Posted 20 September 2014 - 11:50 PM

How about letting Clans drop into a map dispersed and staggered, like at four corners of a map at 30 second intervals, and letting IS start out as a blob?

That way you don't have to change the mechs themselves and still achieve balance. How come I am so smart?

#200 Zuri Prime

    Member

  • PipPipPipPipPip
  • The Raider
  • The Raider
  • 120 posts
  • LocationThe Periphery

Posted 21 September 2014 - 12:18 AM

I'm sure this has been answered, but I can't find it on this forum. My search-fu is weak.

If I'm in an IS-aligned mercenary organization, can I still pilot Clan mechs? Otherwise I'll pass on the new clan packages.





23 user(s) are reading this topic

0 members, 23 guests, 0 anonymous users