Jump to content

Matchmaker Hangs And Wait Times


289 replies to this topic

#181 AlmightyAeng

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 3,905 posts

Posted 30 April 2014 - 07:29 AM

View PostRetroActive, on 30 April 2014 - 07:26 AM, said:


Your scenario assumes the pilots are of equal skill. ZERO pilots are exactly equal in skill. In a typical match you can have a wide variety of skill sets.

Take this scenario. You have a Dragon Slayer on each side with the same exact loadout. The pilot on one side is mediocre and the pilot on the other is elite. How will your BV system balance this?


If you're looking at how 'mechs are balanced, you have to assume outside factors, like skill, are equal...otherwise why are you even discussing this?

Fine...let's get anecdotal then. I've NEVER had a locust kill me while I'm in my Atlas. I've enjoyed hilariously one-shotting them often.

Edited by Ghost Badger, 30 April 2014 - 07:29 AM.


#182 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 30 April 2014 - 07:31 AM

View PostRetroActive, on 30 April 2014 - 07:26 AM, said:


Your scenario assumes the pilots are of equal skill. ZERO pilots are exactly equal in skill. In a typical match you can have a wide variety of skill sets.

When balancing mechs and weapons, you assume equally skilled pilots. Not really sure why such a fundamental aspect of balance needs to be explained.

View PostRetroActive, on 30 April 2014 - 07:26 AM, said:

Take this scenario. You have a Dragon Slayer on each side with the same exact loadout. The pilot on one side is mediocre and the pilot on the other is elite. How will your BV system balance this?

They are the same mech. Their battlevalue is equal.
Battlevalue is for measuring mech value.. it's not a measurement of pilot skill. It's not intended to be a matchmaking system. It's a single metric which would be used within a larger matchmaking system.

#183 Mister Blastman

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 8,444 posts
  • LocationIn my Mech (Atlanta, GA)

Posted 30 April 2014 - 07:39 AM

I don't care about 3/3/3/3. Fix the HIT DETECTION FOR CRYING OUT LOUD!!!!!!!!
Medium wubsteps aren't registering. Nothing is, not with 100% certainty. 30-40% of all weapons aren't registering for me, all the time.

#184 RetroActive

    Member

  • PipPipPipPipPipPip
  • 405 posts
  • LocationFL, USA

Posted 30 April 2014 - 07:40 AM

View PostGhost Badger, on 30 April 2014 - 07:29 AM, said:


If you're looking at how 'mechs are balanced, you have to assume outside factors, like skill, are equal...otherwise why are you even discussing this?

Fine...let's get anecdotal then. I've NEVER had a locust kill me while I'm in my Atlas. I've enjoyed hilariously one-shotting them often.


I don't know why we are using the Atlas vs. Locust scenario. That one is a no brainer. Let's switch to something like Atlas vs. Victor. With the current meta, many would argue that a Victor would have a higher BV than the Atlas. But, based on skill, the Atlas could greatly out perform the Victor.

The point is, no matter what match-making system you use, skill will always trump as long as the weight class is similar.

#185 RetroActive

    Member

  • PipPipPipPipPipPip
  • 405 posts
  • LocationFL, USA

Posted 30 April 2014 - 07:44 AM

View PostRoland, on 30 April 2014 - 07:31 AM, said:


They are the same mech. Their battlevalue is equal.
Battlevalue is for measuring mech value.. it's not a measurement of pilot skill. It's not intended to be a matchmaking system. It's a single metric which would be used within a larger matchmaking system.


Fair enough on the BV being part of a larger system. I just don't see how it's any better than 3/3/3/3, exact weight matching, etc.

#186 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 30 April 2014 - 07:55 AM

View PostRetroActive, on 30 April 2014 - 07:40 AM, said:


I don't know why we are using the Atlas vs. Locust scenario. That one is a no brainer.

The point is that it's a no-brainer, and illustrates an extreme case where one mech is clearly better than the other.

Quote

Let's switch to something like Atlas vs. Victor. With the current meta, many would argue that a Victor would have a higher BV than the Atlas. But, based on skill, the Atlas could greatly out perform the Victor.

You eliminate skill from the equation, because it's irrelevant to the value of the mechs.

Quote

The point is, no matter what match-making system you use, skill will always trump as long as the weight class is similar.

Even if the weight class isn't similar, skill can still trump BV.. but why are you saying that? That's how it should work. Better pilots will make better use of poor equipment. That's fine.

That doesn't mean that BV has no meaning, or shouldn't be considered.

View PostRetroActive, on 30 April 2014 - 07:44 AM, said:


Fair enough on the BV being part of a larger system. I just don't see how it's any better than 3/3/3/3, exact weight matching, etc.

It's better than 3/3/3/3 or weight matching because mech class or weight aren't actually good measures of mech utility.

A locust is not equal to a Jenner, despite them both being the same weight class.

An Awesome is not equal to a Victor, despite them both being the same weight class and the same tonnage.

#187 Cimarb

    Member

  • PipPipPipPipPipPipPipPipPip
  • Caladbolg
  • Caladbolg
  • 3,912 posts
  • Twitter: Link
  • Twitch: Link
  • LocationA hop, skip and jump from Terra

Posted 30 April 2014 - 08:26 AM

View PostRoland, on 30 April 2014 - 07:55 AM, said:

The point is that it's a no-brainer, and illustrates an extreme case where one mech is clearly better than the other.

You eliminate skill from the equation, because it's irrelevant to the value of the mechs.

Even if the weight class isn't similar, skill can still trump BV.. but why are you saying that? That's how it should work. Better pilots will make better use of poor equipment. That's fine.

That doesn't mean that BV has no meaning, or shouldn't be considered.

It's better than 3/3/3/3 or weight matching because mech class or weight aren't actually good measures of mech utility.

A locust is not equal to a Jenner, despite them both being the same weight class.

An Awesome is not equal to a Victor, despite them both being the same weight class and the same tonnage.


While I disagree almost completely with Roland's method of getting the BV, he is completely right about the value of it in matchmaking. Roland's example above about the Awesome and Victor is spot on, as the current matchmaker thinks they are equivalent and we can all agree that is definitely NOT the case.

BV (however you determine it) is the mech portion of the equation, which should be combined with the pilots skill (Elo - the method of which is also up for debate) to give the most accurate "value" to that individual in that mech.

#188 RetroActive

    Member

  • PipPipPipPipPipPip
  • 405 posts
  • LocationFL, USA

Posted 30 April 2014 - 08:29 AM

View PostRoland, on 30 April 2014 - 07:55 AM, said:

The point is that it's a no-brainer, and illustrates an extreme case where one mech is clearly better than the other.


You eliminate skill from the equation, because it's irrelevant to the value of the mechs.


Even if the weight class isn't similar, skill can still trump BV.. but why are you saying that? That's how it should work. Better pilots will make better use of poor equipment. That's fine.

That doesn't mean that BV has no meaning, or shouldn't be considered.


It's better than 3/3/3/3 or weight matching because mech class or weight aren't actually good measures of mech utility.

A locust is not equal to a Jenner, despite them both being the same weight class.

An Awesome is not equal to a Victor, despite them both being the same weight class and the same tonnage.


Note that I said "fair enough" in regard to BV being part of a larger MM system. I was operating under the impression that you were suggesting BV as a magic solution to all MM woes.

All your points make sense, but that doesn't necessarily mean that BV is the best solution. I like the idea of the variety that 3/3/3/3 will bring. With BV, I think you'd still see the assault/heavy fests that we currently see. People like driving Victors, Highlanders, Stalkers and Atlases. BV would still have to match those on the opposite team, assuming they have similar loadouts.

#189 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 30 April 2014 - 08:44 AM

View PostRetroActive, on 30 April 2014 - 08:29 AM, said:


Note that I said "fair enough" in regard to BV being part of a larger MM system. I was operating under the impression that you were suggesting BV as a magic solution to all MM woes.

All your points make sense, but that doesn't necessarily mean that BV is the best solution. I like the idea of the variety that 3/3/3/3 will bring. With BV, I think you'd still see the assault/heavy fests that we currently see. People like driving Victors, Highlanders, Stalkers and Atlases. BV would still have to match those on the opposite team, assuming they have similar loadouts.

Actually, what my system would propose would be a limit on BV, so you wouldn't end up seeing all of the best mechs all the time.

#190 IIIuminaughty

    Member

  • PipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 1,445 posts
  • LocationVirginia

Posted 30 April 2014 - 09:28 AM

View PostRebas Kradd, on 29 April 2014 - 01:10 PM, said:


Not necessarily. Nobody seems to understand how internal testing works. There are factors in live environments that just don't exist in test environments.

Instead of just accusing PGI of incompetence, the community needs to learn a thing or two about software development and accept the fact that Bugs Happen™ no matter how hard you test for them. This is an unfortunate one, but give them a few hours and see what they come up with.

View PostNikolai Lubkiewicz, on 29 April 2014 - 01:25 PM, said:


It is not unusual for changes to be seen in behavior of a new feature between the stages of internal testing with a small team of less than 50, public testing with a few hundred players who are active, willing and able, and live service in which the entire player population is involved. Public Testing helped reveal the issue described here but not the one we face in the OP, which only became a noticeable issue when normal population was factored in.


I Understand what you guys are saying, I really do.
But If it is a public test why not open it up for everyone? Last time I checked you had to sign up for it in a certain amount of time to be eligible for it.
So if everyone was testing it this bug would have been recognized then, since it happened when everyone was on the server.

Not trying to sound like an butt, just trying to understand and give future suggestions, ya digg?

#191 ApolloKaras

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,974 posts
  • LocationSeattle, Washington

Posted 30 April 2014 - 09:31 AM

View PostIIIuminaughty, on 30 April 2014 - 09:28 AM, said:


I Understand what you guys are saying, I really do.
But If it is a public test why not open it up for everyone? Last time I checked you had to sign up for it in a certain amount of time to be eligible for it.
So if everyone was testing it this bug would have been recognized then, since it happened when everyone was on the server.

Not trying to sound like an butt, just trying to understand and give future suggestions, ya digg?


How much more open could it get?? All you had to do was download the client.

You had to be registered and had an account with MWO on a certain date. That date was when the "snapshot" of your account was taken, to be ported over to the test server.

I kept seeing the same folks over and over again, just a lot of the community was a no show... Yet people are blaming PGI to no end it doesn't make sense.

Edit Grammar ^_^

Edited by Saxie, 30 April 2014 - 09:34 AM.


#192 IIIuminaughty

    Member

  • PipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 1,445 posts
  • LocationVirginia

Posted 30 April 2014 - 09:34 AM

View PostSaxie, on 30 April 2014 - 09:31 AM, said:


How much more open could it get?? All you had to do was download the client.

You had to be was registered and had an account with MWO on a certain date. That date was when the "snapshot" of your account was taken, to be ported over to the test server.

I kept seeing the same folks over and over again, just a lot of the community was a no show... Yet people are blaming PGI to no end it doesn't make sense.

Actually It wouldn't allow me on and I downloaded the test server as soon as I saw the announcement come up that day. Soooooooo.....

#193 AdamBaines

    Member

  • PipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 1,384 posts

Posted 30 April 2014 - 09:40 AM

View PostIIIuminaughty, on 30 April 2014 - 09:34 AM, said:

Actually It wouldn't allow me on and I downloaded the test server as soon as I saw the announcement come up that day. Soooooooo.....


What does that mean? Just because you had an issue with getting on, the way it was done is wrong? Id say that either:

A) you downloaded too early. You said as soon as the announcement came up you DLed? Well when the announcement first came up, there was nothing new to DL. The links provided would just reinstall the old stuff. I know this because I was a dummy that did that :-) So you MAY have just Dled the same old version.

B ) Maybe it was just bad luck for you.

PGI was inclusive of everyone as long as you had an account when they did a snapshot of the DB (and that is standard Dev procedures in ANY SDLC). The only issue I had was that I wish they let people know further ahead of time when the Test was happening. 2-3 days notice before was a bit short notice. Even that is not really all that bad.

Edited by AdamBaines, 30 April 2014 - 09:43 AM.


#194 IIIuminaughty

    Member

  • PipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 1,445 posts
  • LocationVirginia

Posted 30 April 2014 - 10:05 AM

They also have it running during the middle of the day, most people work. during the middle of the day. Why not have it up during prime time. I also know that the devs have to go home sooner or later but, we are leaving feed back in the forums.

#195 Gil Wallace

    Rookie

  • 1 posts

Posted 30 April 2014 - 10:10 AM

View PostSaxie, on 30 April 2014 - 09:31 AM, said:



I kept seeing the same folks over and over again, just a lot of the community was a no show... Yet people are blaming PGI to no end it doesn't make sense.

Edit Grammar :D


If there was lack of participation to a mass test then its not a mass test. That means you dont have proper testing done. If you dont have proper testing done, you darn well do not roll forward with any modifications. That is development basics or 101. Now with all that said PGI can be blamed because they could have held off on the roll out and made sure more testing was completed and made sure they had a true mass test. An held off any modifications until they did have a true mass test.

#196 Silverlance

    Member

  • PipPipPipPipPip
  • The Merciless
  • The Merciless
  • 178 posts
  • Facebook: Link
  • Twitter: Link
  • LocationThe Periphery, one planet at a time.

Posted 30 April 2014 - 10:16 AM

You know. Every forum that I browse here. A lot of you guys have got some fantastic ideas and major kudos to you all. Really and sincerely.
But the one thing you unfortunately all have to remember is this:
This is PGI. They won't listen and they'll never take ideas from you because then they'd have to credit &/or pay you. Or they're just plain old ********.
They had a chance to listen and implement so many times and haven't and went their own ways and they're still failing at such simple things. Keep in mind, this is a game developer who's game couldn't even hack it on Facebook. ^_~
Maybe one day they'll actually open up the floor to implement actual user ideas to make this game better than the joke metafest it currently is. When I signed up as a Legendary Founder for Mechwarrior, I didn't think I was buying into another version of Call of Duty.

#197 Tannhauser Gate

    Member

  • PipPipPipPipPipPipPipPip
  • The Blood-Eye
  • The Blood-Eye
  • 1,302 posts
  • LocationAttack ship off the Shoulder of Orion

Posted 30 April 2014 - 10:28 AM

View PostGil Wallace, on 30 April 2014 - 10:10 AM, said:


If there was lack of participation to a mass test then its not a mass test. That means you dont have proper testing done. If you dont have proper testing done, you darn well do not roll forward with any modifications. That is development basics or 101. Now with all that said PGI can be blamed because they could have held off on the roll out and made sure more testing was completed and made sure they had a true mass test. An held off any modifications until they did have a true mass test.


I would agree with you if this was a traditionally developed game by a large game developer. We would expect fully vetted, fully tested content with a low amount of bugs. But, PGI is not EA with 1500+ compensated game testers. We are the beta testers. Waiting to fully test all patches and updates would mean months of extra delays.

I totally expect content to be buggy and need balancing and patching. That was the planned dev model for MWO.

#198 Name140704

    Member

  • PipPipPipPipPipPipPipPip
  • 1,196 posts
  • LocationBehind You

Posted 30 April 2014 - 10:38 AM

View PostLakeDaemon, on 30 April 2014 - 10:28 AM, said:


I would agree with you if this was a traditionally developed game by a large game developer. We would expect fully vetted, fully tested content with a low amount of bugs. But, PGI is not EA with 1500+ compensated game testers. We are the beta testers. Waiting to fully test all patches and updates would mean months of extra delays.

I totally expect content to be buggy and need balancing and patching. That was the planned dev model for MWO.



Beta ended eptember 17th, 2013. U.I. 2.0 soon to follow, with Community Warfare 90 days after.

Don't you remember?

#199 Osric Lancaster

    Member

  • PipPipPipPipPipPip
  • 447 posts

Posted 30 April 2014 - 10:42 AM

View PostSilverlance, on 30 April 2014 - 10:16 AM, said:

. . . because then they'd have to credit &/or pay you.


What? No they wouldn't. They actually fixed NARC while referencing a forum thread as the driving force behind the change. They started dividing 'Mech crotch hit-boxes between the legs because of a big effort post. Those were good things, and I wish there was more of it, but PGI is in no way required to credit people for giving them ideas. It'd be nice of them to acknowledge things with a 'yeah this is a good idea' or 'no, bad idea, here's why' but . . .

Seriously though, the best things to have happened to MWO in the recent past have come from the rare occasions where PGI listened. Can you imagine how amazing it would be if they actually looked at the forums and talked with us about game play? Or talked about things they were thinking of doing before they did them and then listened to feedback? Even for someone without the authority to make global game changes to just give an opinion on something would be amazing because we'd know you're interested in the game. Like -

Hey Niko, what's your opinion on the LB 10-X or SRMs as a weapon? How do you feel about the critical system? What's your take on . . . hell, anything. What part of MWO do you like? What do you think could be improved?

Edited by Osric Lancaster, 30 April 2014 - 11:13 AM.


#200 Morbid Jester

    Member

  • Pip
  • Overlord
  • Overlord
  • 18 posts
  • Location, Location, Location

Posted 30 April 2014 - 10:43 AM

On a lighter note:

Game mode selection in public matches works for me now... n=1 says I drop faster - if I do not hang in limbo





4 user(s) are reading this topic

0 members, 4 guests, 0 anonymous users