Jump to content

Could We Try A Week/patch-Cycle With Elo Turned Off, Please?


134 replies to this topic

#121 The Amazing Atomic Spaniel

    Member

  • PipPipPipPipPipPipPip
  • 932 posts
  • LocationBath, UK

Posted 27 October 2013 - 10:34 AM

View Postaseth, on 27 October 2013 - 09:18 AM, said:

The tonnage matching these days is pretty terrible. When I first started playing, it was much, much better - in most cases the number of lights, mediums, heavies, and assaults on each team were pretty similar.

It was vastly superior to our current matchmaking.


Earlier versions of the MM did indeed balance purely by weight class, so for every assault mech on one team the MM would assign an assault mech on the other. Unfortunately, it considered nothing else and so premades on comms who had been playing for months would be matched up against random PUGs, including those in their first few games. As a result, everyone ended up with frequently unsatisfying matches. The experienced players faced opponents who were no challenge at all and so ended up with 10:1 K/D and W/L ratios and the poor newcomers were massacred match after match. Some newcomers will have persevered but many will have written off MWO as a lost cause and gone on to other games.

Almost as bad was the use of weight *classes*, rather than tonnages, in the forming of teams. I really wanted to play Awesomes (for sentimental reasons), but every time I took an Awesome I was probably giving the other team an Atlas. There were games when I would look around after spawning and see *all* our assaults were Awesomes - at which point I knew we had already lost because we stood little chance against an equal number of Atlases.

I guess the current MM attempts to balance by skill (Elo) alone and it just hopes that the tonnages will balance out now that we are at 12 vs 12. It sometimes gives odd results, but it's much better than the previous versions. I'm a very plodding player, but my W/L has moved to be almost exactly 1/1, whereas when I started it was much, much worse.

#122 Deathlike

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • Littlest Helper
  • Littlest Helper
  • 29,240 posts
  • Location#NOToTaterBalance #BadBalanceOverlordIsBad

Posted 27 October 2013 - 10:43 AM

View PostRocketDog, on 27 October 2013 - 10:34 AM, said:

Earlier versions of the MM did indeed balance purely by weight class, so for every assault mech on one team the MM would assign an assault mech on the other. Unfortunately, it considered nothing else and so premades on comms who had been playing for months would be matched up against random PUGs, including those in their first few games. As a result, everyone ended up with frequently unsatisfying matches. The experienced players faced opponents who were no challenge at all and so ended up with 10:1 K/D and W/L ratios and the poor newcomers were massacred match after match. Some newcomers will have persevered but many will have written off MWO as a lost cause and gone on to other games.


Sad truth.

Quote

Almost as bad was the use of weight *classes*, rather than tonnages, in the forming of teams. I really wanted to play Awesomes (for sentimental reasons), but every time I took an Awesome I was probably giving the other team an Atlas. There were games when I would look around after spawning and see *all* our assaults were Awesomes - at which point I knew we had already lost because we stood little chance against an equal number of Atlases.


Weight based matching did have flaws, and that was the extreme part of it. Back when the Raven-3L ws the goto light, any other light outside of the Jenner being fielded was a wasted slot.

Running a Cicada during that time was done to minimize the 3L on the other end, while still being able to field ECM (outside of an Atlas for Assaults). At least you could face Hunchies and Cents, and not be overly threatening to a Cicada.

Awesomes... you were just wasting a slot that could have been filled with a Stalker (which is still true today).

#123 Alondo

    Member

  • PipPipPip
  • Veteran Founder
  • Veteran Founder
  • 85 posts
  • LocationKansas, USA

Posted 27 October 2013 - 01:22 PM

I'm down with the sickness. Let's give it a try. :huh:

#124 Iskareot

    Member

  • PipPipPipPipPipPip
  • The Universe
  • The Universe
  • 433 posts
  • Google+: Link
  • LocationNW,IN

Posted 28 October 2013 - 06:03 AM

View PostKunae, on 22 October 2013 - 05:36 AM, said:

Assertion: The Elo system is very flawed, polluted with stats from both grouping and solo, and with their mechanic of balancing high-Elo with a proportion of very low Elo, in matches, to "even things out", it hurts the game more than it helps.

I would suggest that PGI turns off Elo for one week or patch-cycle so everyone can see how it plays. Just use a rough weight-class matching, and let it go at that.

It is my belief that the resultant matches will be the same or better than ones using Elo. I would just like to be able to test this theory.



Yes and OR simply give a choice at search for game launch. Also add in premade allowed or pug and then you will see some interesting numbers and or posts lol.

#125 Iskareot

    Member

  • PipPipPipPipPipPip
  • The Universe
  • The Universe
  • 433 posts
  • Google+: Link
  • LocationNW,IN

Posted 28 October 2013 - 06:07 AM

View PostRocketDog, on 27 October 2013 - 10:34 AM, said:


Earlier versions of the MM did indeed balance purely by weight class, so for every assault mech on one team the MM would assign an assault mech on the other. Unfortunately, it considered nothing else and so premades on comms who had been playing for months would be matched up against random PUGs, including those in their first few games. As a result, everyone ended up with frequently unsatisfying matches. The experienced players faced opponents who were no challenge at all and so ended up with 10:1 K/D and W/L ratios and the poor newcomers were massacred match after match. Some newcomers will have persevered but many will have written off MWO as a lost cause and gone on to other games.

Almost as bad was the use of weight *classes*, rather than tonnages, in the forming of teams. I really wanted to play Awesomes (for sentimental reasons), but every time I took an Awesome I was probably giving the other team an Atlas. There were games when I would look around after spawning and see *all* our assaults were Awesomes - at which point I knew we had already lost because we stood little chance against an equal number of Atlases.

I guess the current MM attempts to balance by skill (Elo) alone and it just hopes that the tonnages will balance out now that we are at 12 vs 12. It sometimes gives odd results, but it's much better than the previous versions. I'm a very plodding player, but my W/L has moved to be almost exactly 1/1, whereas when I started it was much, much worse.



AND this is why to an exact point of why we need choice now. At one time in beta, no fine.. I understand.. now that it is live - CHOICE would be ideal. I would think it would eliminate a lot of issues. I think having some choices would also curb the complaints. It gets old seeing the imbalance you see now... and you see it over 50% of the time easy. I have ran into some matches that have just been a waste of time and caused me to close the game due to how failed it was.

If they want people to keep spending money they better start giving a choice to have a better chance, or balance to drops thats for sure.

#126 Roland

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 8,260 posts

Posted 28 October 2013 - 06:11 AM

Quote

Earlier versions of the MM did indeed balance purely by weight class, so for every assault mech on one team the MM would assign an assault mech on the other. Unfortunately, it considered nothing else and so premades on comms who had been playing for months would be matched up against random PUGs, including those in their first few games.

The thing is though, that if you lose a match because you are worse, so be it. That's fair. That's how games work.

But when you lose a match because the other team out-tonned you, then it's garbage. It's inherently unfair, and it's that type of thing which makes folks stop playing a game.

When you lose because you got beat, you have the opportunity to get better. When you lose because the matchmaker screwed you in terms of tonnage, then you have no chance to improve your game. You just get screwed.

#127 Asyres

    Member

  • PipPipPipPipPipPip
  • 433 posts

Posted 28 October 2013 - 06:17 AM

I actually stopped playing this game for several months before ELO was implemented. As flawed as the current system is, it's head and shoulders above what preceded it.

You want a week with mostly horrible matches? Fine. May I recommend the week of Nov 10th? I could stand a reason not to play during finals week.

#128 Kunae

    Member

  • PipPipPipPipPipPipPipPipPip
  • 4,303 posts

Posted 28 October 2013 - 07:50 AM

View PostRoland, on 28 October 2013 - 06:11 AM, said:

The thing is though, that if you lose a match because you are worse, so be it. That's fair. That's how games work.

But when you lose a match because the other team out-tonned you, then it's garbage. It's inherently unfair, and it's that type of thing which makes folks stop playing a game.

When you lose because you got beat, you have the opportunity to get better. When you lose because the matchmaker screwed you in terms of tonnage, then you have no chance to improve your game. You just get screwed.

While the weight thing is an issue... when you lose because you're teamed with people who don't understand that LRMs have a 180m min range, and that 1 SRM2 and 1 Med laser is not a good build for a thunderbolt(True story, I actually saw this), or can't figure out that they shouldn't try exchanging fire at 600m with an enemy PPC boat, when they've only got 1 LL and the rest ML, is not a good plan... it makes for not enjoyable losses.

Edited by Kunae, 28 October 2013 - 07:51 AM.


#129 Wispsy

    Member

  • PipPipPipPipPipPipPipPipPip
  • Talon
  • Talon
  • 2,007 posts

Posted 28 October 2013 - 08:27 AM

View PostKunae, on 28 October 2013 - 07:50 AM, said:

While the weight thing is an issue... when you lose because you're teamed with people who don't understand that LRMs have a 180m min range, and that 1 SRM2 and 1 Med laser is not a good build for a thunderbolt(True story, I actually saw this), or can't figure out that they shouldn't try exchanging fire at 600m with an enemy PPC boat, when they've only got 1 LL and the rest ML, is not a good plan... it makes for not enjoyable losses.


I only very rare ever see things like that except very occasionally at that weird time of day where you spend 30 mins waiting for match and win 40 games in a row if you do actually get one...


During na peak time however You can tell Elo is doing its job. All the faces are the same and it is very rare to see any lights or mediums as everybody can aim well.

#130 Kunae

    Member

  • PipPipPipPipPipPipPipPipPip
  • 4,303 posts

Posted 28 October 2013 - 08:28 AM

View PostWispsy, on 28 October 2013 - 08:27 AM, said:

I only very rare ever see things like that except very occasionally at that weird time of day where you spend 30 mins waiting for match and win 40 games in a row if you do actually get one...


During na peak time however You can tell Elo is doing its job. All the faces are the same and it is very rare to see any lights or mediums as everybody can aim well.

Are you dropping solo at all, lately?

#131 Wispsy

    Member

  • PipPipPipPipPipPipPipPipPip
  • Talon
  • Talon
  • 2,007 posts

Posted 28 October 2013 - 08:36 AM

View PostKunae, on 28 October 2013 - 08:28 AM, said:

Are you dropping solo at all, lately?


Not in a light but I do in everything else. Can't drop solo light with so many ssrms around, guaranteed damage on low armour targets lol...need somebody who can protect me from them.

Edit: well unless it is peak time, then people generally know what to do and if someone charges at me they just get shredded, however my friends apparently feel emotionally wounded when I do this so I mostly play with them instead.

Edited by Wispsy, 28 October 2013 - 08:38 AM.


#132 Kunae

    Member

  • PipPipPipPipPipPipPipPipPip
  • 4,303 posts

Posted 28 October 2013 - 08:40 AM

View PostWispsy, on 28 October 2013 - 08:36 AM, said:

Not in a light but I do in everything else. Can't drop solo light with so many ssrms around, guaranteed damage on low armour targets lol...need somebody who can protect me from them.

Edit: well unless it is peak time, then people generally know what to do and if someone charges at me they just get shredded, however my friends apparently feel emotionally wounded when I do this so I mostly play with them instead.

Oh come now... you know you're just supposed to run away from those SSRMs... right? :blink:

It's good of you to emotionally support your fragile friends. Very magnanimous of you. :)

#133 Iskareot

    Member

  • PipPipPipPipPipPip
  • The Universe
  • The Universe
  • 433 posts
  • Google+: Link
  • LocationNW,IN

Posted 28 October 2013 - 09:08 AM

View PostMischiefSC, on 25 October 2013 - 09:06 PM, said:

Splitting group from solo Elo isn't a bad idea at all.

@Kunae;

The problem is that your experience is anecdotal and your stats distinctly above average. If you've got a couple hundred/few hundred drops under you and you mostly do lights then statistically you're going to be a ways up in Elo. Hence your experience is going to differ from most peoples.

The point of Elo is to do two things:

1. Try to get a more consistent difficulty, statistically, for every player in every match.

2. To protect people with a 1.0 win/loss and 1.0 KDR (or less) from players like you.

It does that. Any sort of win/loss over a 1.5 after a good 300 or so matches means you're doing very well and see only a small sliver of the more average part of the player base. That's good, that's Elo doing what it's supposed to.

Removing Elo will NOT result in you seeing less noobs - it will result in you seeing more.

Does that make sense? The only reason someone with a good win/loss and good KDR would want to play without Elo would be access to more low Elo players since people with equal or better Elo than yourself will be statistically far, far less common than you see now. I don't think that's what you want. It sounds like you're just frustrated that you're seeing people on your team who are.... struggling, and now you feel like it's *intentional*. That it's not just bad luck but like the system is hamstringing you by dropping you with mouth-breathing window-lickers who LOVE 3PV and feel like it's the best mode to fight in.

What's important to understand is that you're seeing far, far less of them than you would without Elo. Elo is protecting you from them (well, them from you really) to a huge degree. If MW:O has a population anything like most MMO-style games (and I've worked for a few) then in any given hour ~20% of the active playerbase is online. Your top 10-20% skill-wise of players actually populate ~1% of the total population. So ~1/5th of 1% of people playing in any given hour are excellent players, maybe as much as 10 or 20% are 'good', the remaining 80-90% are pretty average. As in win/loss of 1.0-1.2 and a KDR of 0.8 to 1.2. This means that in any given game of the 24 players if you, Kunae, were dropping solo in pugs there might be one or at most 2 people in the whole match are are even approximately your own skill. The remaining 21 players would be the window-lickers you're trying to avoid. While your KDR would skyrocket you would curb-stomp newer players ever game, forcing the average win/loss and KDR of the rest of the pbase down. Is that what you really want? I don't think it's what everyone else wants.

Here's another thing to keep in mind. You get a little better every game you play. For you to have kept a consistent win/loss and KDR several months after the release of Elo means that you are, at this point, considerably better than you were pre-Elo. The caliber of the people you're playing against is better you're just improving at about the same rate that your Elo is rising. That's awesome! Eventually you'll start to even out but at this point it's not a good experience for everyone else to drop you in the game without Elo. It's not fair and it's not fun for everyone else. Hopefully not for you either.

Elo has plenty of room to grow and improve. Nobody questions that. Without it however it would be back to 10% of the games population ROFLSTOMPing the remaining 90%, only worse. New players are as good (bad) as they ever were. We've all gotten considerably better because we're playing better opponents. Perhaps the problem is your expectations are higher now? Don't you remember when it was a great game if over 1/2 of your team just went the same direction? Now it's only 1 or at most 2 out of 12 that tend to pull a goober.

Splitting group from solo Elo is good though. That's a place I think we could tighten the experience up for everyone.




Oh how I love logic and actual information being used. Sadly this is the issue I discuss constantly. What is seen by each person in some respects is different due to the ELO system and MM with the combo of Pug vs Premades.

There is no way to clear this up other then be given a choice. Do the math... Check box if you do not mind being with premades in your drop - do not check if you rather wait for a filled pug group. Offer 1% more cbills in the premade ones because lets face it as much as the constant premade groups WANT to think that people just love getting owned by a smarter group that is not of pug mentality. (As in you are in a group that will listen and do what is asked) vs the Pug group that has two guys driving around in circles -- one thinking he is rambo and one thinking it is ok to stand at the edge of the map alone by himself waiting for someone to go by... BUT hey... Im sure that KDR you gain from it should help you due to that great MM system.

If we have options I am thinking this will all change peoples KDR's for sure.

#134 Kunae

    Member

  • PipPipPipPipPipPipPipPipPip
  • 4,303 posts

Posted 28 October 2013 - 09:46 AM

View PostIskareot, on 28 October 2013 - 09:08 AM, said:




Oh how I love logic and actual information being used. Sadly this is the issue I discuss constantly. What is seen by each person in some respects is different due to the ELO system and MM with the combo of Pug vs Premades.

There is no way to clear this up other then be given a choice. Do the math... Check box if you do not mind being with premades in your drop - do not check if you rather wait for a filled pug group. Offer 1% more cbills in the premade ones because lets face it as much as the constant premade groups WANT to think that people just love getting owned by a smarter group that is not of pug mentality. (As in you are in a group that will listen and do what is asked) vs the Pug group that has two guys driving around in circles -- one thinking he is rambo and one thinking it is ok to stand at the edge of the map alone by himself waiting for someone to go by... BUT hey... Im sure that KDR you gain from it should help you due to that great MM system.

If we have options I am thinking this will all change peoples KDR's for sure.

This is not about "premades vs pugs". This is about the MM forcing a drastic Elo disparity on one side vs roughly equal on the other.

Clue for you: There are some really really crappy "premades" out there. Just because people are dropping together doesn't mean they're not wearing helmets and drooling on themselves.

#135 DocBach

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 4,828 posts
  • LocationSouthern Oregon

Posted 28 October 2013 - 09:51 AM

View PostKunae, on 28 October 2013 - 09:46 AM, said:


Clue for you: There are some really really crappy "premades" out there. Just because people are dropping together doesn't mean they're not wearing helmets and drooling on themselves.


The helmet is for immersion purposes, dammit!





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users