Jump to content

Elo Is The Word


128 replies to this topic

Poll: ELO is he all that bad? (266 member(s) have cast votes)

is elo ruining you gaming experience?

  1. yes (57 votes [21.43%])

    Percentage of vote: 21.43%

  2. no (180 votes [67.67%])

    Percentage of vote: 67.67%

  3. haven't noticed really (29 votes [10.90%])

    Percentage of vote: 10.90%

is elo fair?

  1. yes (103 votes [38.72%])

    Percentage of vote: 38.72%

  2. no (89 votes [33.46%])

    Percentage of vote: 33.46%

  3. undecided (74 votes [27.82%])

    Percentage of vote: 27.82%

if answered no then how should it be developed?

  1. it needs more work regarding tonnage matching (89 votes [33.46%])

    Percentage of vote: 33.46%

  2. needs more work regarding BVs (28 votes [10.53%])

    Percentage of vote: 10.53%

  3. Holy expletative batman this elo was never needed and simply has to go! (19 votes [7.14%])

    Percentage of vote: 7.14%

  4. um... i voted yes elo is working as intended (95 votes [35.71%])

    Percentage of vote: 35.71%

  5. i'm as of yet not sure what should/could be done (35 votes [13.16%])

    Percentage of vote: 13.16%

Vote Guests cannot vote

#81 Nightfire

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 226 posts
  • LocationAustralia

Posted 15 March 2013 - 08:20 AM

View PostFerretGR, on 15 March 2013 - 08:07 AM, said:


Basing your opinion of a player's skill and whether or not he's a scrub on a single boneheaded move (or even a series of boneheaded moves) is just as bad as basing it on a single bad game, IMHO.

Maybe he's awesome, but tonight, he's drunk. Happens to the best of us... :ph34r:

I would say "boneheaded" only holds up when you look at single instances. I can see the players in my Elo bracket who play regularly in my time slot. I recognize their names and I can gauge their ability in mechs over a larger sample size of matches and as such, their respective threat level. Scientifically speaking, you are correct that this is a far better method for determining a player's skill.

However when I see a name I don't recognize AND he does something, not just strange but what I would call from questionable to outright asinine, you have several ways to go with this. You could say to yourself, "well, I don't have enough data yet and that could have been an honest mistake" or as I often do say, "that was a stupid thing to do and until he demonstrates better tactical reasoning I'll rate him a scrub and a liability to my survival". A lot of these I don't see again so they remain, as you called them, "scrubs".

If we go on to win, well my enjoyment factor really doesn't take a hit so my pre-made and I are happy to laugh at his misfortune and move on. If we lose however, it becomes a bit of a craw that we really could have performed better as a team if we had someone else. If we lose and we had 4 of these players, then it becomes downright frustrating.

So was the player really a scrub or were they just having a bad match? We'll never know for sure but when this pattern happens frequently enough to be recognizable, then you have to admit that it is a reasonable theory to put forward.

#82 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 15 March 2013 - 08:21 AM

View PostMischiefSC, on 15 March 2013 - 07:54 AM, said:

I so predicted this. Even my timing was correct. About few weeks after Elo starts we get the tears over people missing pugstomping and trying to couch it as something else. That 'something else' is 'the rest of my team sucks'.

GLORIOUS.

Was looking back over stuff we discussed a month ago. Here, I'mma quote myself:



So what we have now are people used to pugstomping dropping with teams of people who, honestly? Are probably their exact same skill level but are used to being able to pug without teamwork because the other side always sucked. Now they're losing at LEAST 50% of the time and are used to winning 80%.

How about someone talking about how terrible Elo is post a screenie of their stats? Are they at 0.80 win rate? 0.30? Or are they around 1.0 still and just hating that and now looking for someone to blame?



Wrong. I'm annoyed because the SOLO game went to ****.

4 mans are basically unchanged, and elo has next to no effect on 8 mans because there isn't the population.

#83 MischiefSC

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • The Benefactor
  • The Benefactor
  • 16,697 posts

Posted 15 March 2013 - 08:25 AM

View PostNightfire, on 15 March 2013 - 08:01 AM, said:

MischiefSC, you haven't a clue. Furthermore you're projecting and you're a hatemonger. You add nothing useful, project malice on to others in your chosen group of those you find acceptable to despise.

This is a valid problem, just because is doesn't affect you doesn't mean that it is said to deflect inability to play or inflated stats due to what you perceived as pervasive pug stomping. My win/loss is well above 50% so it isn't sour grapes that I'm losing all the time. What is a problem for me is my enjoyment level when I'm expected to carry those who play like they have no idea., especially when my teams loses as a result.

If you can't put forward a view without casting aspersions upon others, the mods really should just ban you.


Wow, ad hominem much?

Look. What you're doing is effectively saying that when you lose it's because your team is terrible. Do you understand both how that sounds and statistically what that's saying? Your argument is based around the premise that the matchmaker treats you differently than everyone else. Do you understand as well how that comes across?

Your win/loss is still over 50% so you're climbing in Elo and will slowly but consistently be put with better and better people, so again, working as intended. Currently you've put forward a conjecture and a complaint when to all other respective points Elo is working as intended without putting forward anything but unsupported anecdotal evidence. Weight matchmaking is still too lose and PGI has said so but it's being tightened slowly. The problem with matching both weight and Elo is delays in finding matches on both ends of the spectrum and finding a good balance.

Again, your teammates may not be lower in Elo than you - they may just not be good at teamwork i pugs. What Elo isn't doing is just dropping you with a bunch of incompetent boobs to try and force you to lose. The reality is that they're going to be pretty close to your own Elo level. Maybe they're all the ex-pugstompers paying their painful dues.

What you are not doing is having a vastly different experience than everyone else in this game. Sometimes the team I drop with doesn't coordinate. Sometimes even excellent players get dropped by focused fire when the other team really brought their A game.

How many matches have you had since Elo was fully implemented? Dozens? Hundreds? If you're still gaining in Elo then you haven't settled yet and you'll continue to drop with progressively better and better people. Again. This isn't a problem, it's working as intended.

View PostYokaiko, on 15 March 2013 - 08:21 AM, said:



Wrong. I'm annoyed because the SOLO game went to ****.

4 mans are basically unchanged, and elo has next to no effect on 8 mans because there isn't the population.


If your solo game is bad the only question is are you winning or losing more than 50%? If you're winning more you'll continue dropping with better people. Are you saying that somehow you're consistently dropping with worse people now than when prior to Elo everyone you dropped with was completely random?

#84 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 15 March 2013 - 08:28 AM

View PostMischiefSC, on 15 March 2013 - 08:25 AM, said:


Wow, ad hominem much?

Look. What you're doing is effectively saying that when you lose it's because your team is terrible. Do you understand both how that sounds and statistically what that's saying? Your argument is based around the premise that the matchmaker treats you differently than everyone else. Do you understand as well how that comes across?

Your win/loss is still over 50% so you're climbing in Elo and will slowly but consistently be put with better and better people, so again, working as intended. Currently you've put forward a conjecture and a complaint when to all other respective points Elo is working as intended without putting forward anything but unsupported anecdotal evidence. Weight matchmaking is still too lose and PGI has said so but it's being tightened slowly. The problem with matching both weight and Elo is delays in finding matches on both ends of the spectrum and finding a good balance.

Again, your teammates may not be lower in Elo than you - they may just not be good at teamwork i pugs. What Elo isn't doing is just dropping you with a bunch of incompetent boobs to try and force you to lose. The reality is that they're going to be pretty close to your own Elo level. Maybe they're all the ex-pugstompers paying their painful dues.

What you are not doing is having a vastly different experience than everyone else in this game. Sometimes the team I drop with doesn't coordinate. Sometimes even excellent players get dropped by focused fire when the other team really brought their A game.

How many matches have you had since Elo was fully implemented? Dozens? Hundreds? If you're still gaining in Elo then you haven't settled yet and you'll continue to drop with progressively better and better people. Again. This isn't a problem, it's working as intended.



If your solo game is bad the only question is are you winning or losing more than 50%? If you're winning more you'll continue dropping with better people. Are you saying that somehow you're consistently dropping with worse people now than when prior to Elo everyone you dropped with was completely random?



If half of my team is sub 100 damage and there are only two kills from the people that didn't die before doing....well two volleys.... its safe to say that someone isn't carrying their weight.

I saw that a minimum of 4 times yesterday. 5

#85 FerretGR

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,445 posts

Posted 15 March 2013 - 08:29 AM

View PostNightfire, on 15 March 2013 - 08:20 AM, said:

So was the player really a scrub or were they just having a bad match? We'll never know for sure but when this pattern happens frequently enough to be recognizable, then you have to admit that it is a reasonable theory to put forward.


Well, maybe. All I can do is offer anecdotal evidence that might make you consider holding off on forming a conclusion. I'm not going to toot my own horn; I'm okay at this game but I'm hardly top-tier. I recognize some names from these forums when I play. A certain someone whose stats I've seen and who is "good" as far as I can tell has dropped with me exactly twice that I can remember. In the first game, I got LURMED to death stupidly and did like 60 damage (I recognize I made a boneheaded move that lead to this); in the second, I got headshotted and did only 5 (blush) damage before doing so. I fear that that skilled player (who shall remain unnamed, haha) definitely classifies me as a scrub after those two achievements in suckage, but honestly, that level of perfomance is not my norm. I blame the tournament... it was the first one, and I had been playing for hours... I zoned out :ph34r:

Given this, I tend to give folks the benefit of the doubt :P

Edited by FerretGR, 15 March 2013 - 08:31 AM.


#86 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 15 March 2013 - 08:31 AM

View PostFerretGR, on 15 March 2013 - 08:29 AM, said:


Well, maybe. All I can do is offer anecdotal evidence that might make you consider holding off on forming a conclusion. I'm not going to toot my own horn; I'm okay at this game but I'm hardly top-tier. I recognize some names from these forums when I play. A certain someone whose stats I've seen and who is "good" as far as I can tell has dropped with me exactly twice that I can remember. In the first game, I got LURMED to death stupidly and did like 60 damage; in the second, I got headshotted and did only 5 (blush) damage before doing so. I fear that that skilled player (who shall remain unnamed, haha) definitely classifies me as a scrub after those two achievements in suckage, but honestly, that level of perfomance is not my norm. I blame the tournament... it was the first one, and I had been playing for hours... I zoned out :ph34r:

Given this, I tend to give folks the benefit of the doubt :P



Hell the one tournament I killed Broceratops at least 5 times. I didn't do the one last weekend, the times were terrible for me and it was just lame mechanically anyway.

Edited by Yokaiko, 15 March 2013 - 08:31 AM.


#87 Schrottfrosch

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 253 posts

Posted 15 March 2013 - 08:32 AM

How can you even tell that ELO works?

I am being thrown into totally random matches with extreme weight differences - I am talking about teams that are over 100 tons lighter or heavier than my team. That is not even ballancing - it is simply unfair!

Also maybe count in the number of available ECMs - because a team with no ECM against a team with 4 ECMs is unfair, too.

Ok, I know how to circumvent this - Only play Atlas DDC yourself or a Raven 3L - and make your team, if you are in one, to do this, too. But that is barely fun, is it?

#88 FerretGR

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,445 posts

Posted 15 March 2013 - 08:36 AM

I never run ECM. I don't even own a 3L OR a DDC. I'm making out okay.

Edited by FerretGR, 15 March 2013 - 08:36 AM.


#89 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 15 March 2013 - 08:38 AM

View PostFerretGR, on 15 March 2013 - 08:36 AM, said:

I never run ECM. I don't even own a 3L OR a DDC. I'm making out okay.



You would be shocked how many people miss the hostile ECM symbol coming up.....

#90 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 15 March 2013 - 08:47 AM

View PostThontor, on 15 March 2013 - 08:38 AM, said:

Are you saying you have never been one of those people who does early with less than 100 damage done? Because I regularly place in the top 2-4 damage done as well... Averaging at least 300... But sometimes I happen to be the guy the enemy targets and focuses fire on first...

I have said it before, but just because someone does early with little damage done doesn't make them a worse player than you, it happens to all of us... If you come back and tell me you never die early with under 100 damage done, I call BS


Never, absolutely not. I drive lights far and away more than any other weight recently, mainly because of the weight disparities.

That lends itself to those games, however four people dying sub-100 damage in heavies and assault isn't bad luck, nor isolated.
Most played mechs
Raven 53
K2 24
D-DC 20 (probably 8 man night)
Flame 19

The Raven ensures that I (usually) don't get LRM'd into the floor because the D-DC is running counter....for a raven (this actually happened yesterday vs 4 LRM boats) and that we have a scout, because getting suddenly atlased sucks.


Also looking I've only played 160 matches since the last patch, that is about a third of my usual, and I didn't have anything to do but work this week, and a couple extra days off.

Edited by Yokaiko, 15 March 2013 - 08:51 AM.


#91 MischiefSC

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • The Benefactor
  • The Benefactor
  • 16,697 posts

Posted 15 March 2013 - 08:50 AM

View PostYokaiko, on 15 March 2013 - 08:28 AM, said:



If half of my team is sub 100 damage and there are only two kills from the people that didn't die before doing....well two volleys.... its safe to say that someone isn't carrying their weight.

I saw that a minimum of 4 times yesterday. 5


Here's the thing though. That happens to everyone, especially at high Elo - it's no so much that your team sucked but that your team was just not coordinated and the other team was. THAT is what turns high Elo matches into rolls - not that your team has no clue how to play but that they didn't coordinate and the other team did.

Thontor brings up a great point - you've never been killed with less than several hundred damage? Ever? You're never not the top of the list?

For the arguments that you and Nightfire are making to hold statistical significance you guys would have to NEVER have a bad game. Ever. You play perfect every game and are always the best performer every time you drop.

Do you see how that comes across?

What is more likely is that you're all speaking to a form of cognitive bias like availability heuristic. You're focusing on the times other people have done poorly and you have done well and it's skewing your statistical measurement. If your win/loss is over 50% then actually most the time your team is doing great. The other team is probably looking a lot like your team does when you lose - your team gels, focuses well and coordinates and the other team dies before they even have a chance to do much.

When this happens to you the most significant examples will be the ones you remember best - the ones where you did well and your team did poorly.

The statistical reality however is that you're winning over 50%, that means that most of the time your team is doing great. Statistically there will be times where your team did well and you didn't so more to the point your teammates are probably kicking a$$ a good 60% of the time. Sometimes the other team is just better though.

High Elo matches are going to involve a lot of very good players. It's not going to pull in noobs. The problem with high Elo players though is that when they're good and they coordinate they're freaking amazing. When they're good and they don't, then they're just good. One side coordinates well, otherside doesn't, you're dead before you even had a chance.

Elo. Working as intended. The statistical probability of you dropping with people who suck is so low as to be almost insignificant. More likely you're dropping at times when game populations are low and you're seeing a wider swing though and this makes the relative win/loss swing wider.

You are however have the exact same game experience as everyone else.

#92 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 15 March 2013 - 08:55 AM

View PostMischiefSC, on 15 March 2013 - 08:50 AM, said:

Elo. Working as intended. The statistical probability of you dropping with people who suck is so low as to be almost insignificant. More likely you're dropping at times when game populations are low and you're seeing a wider swing though and this makes the relative win/loss swing wider.

You are however have the exact same game experience as everyone else.



Then that will be the end of me playing, because its no longer fun.

View PostThontor, on 15 March 2013 - 07:59 AM, said:

Are you telling me that you have never once just been killed quickly before you had a chance to do much damage? Just because the enemy decided you were their first target?



Why where you out in the open?

That is mechwarrior 101.....don't be very far from cover.

#93 Mackman

    Member

  • PipPipPipPipPipPipPip
  • 746 posts
  • LocationCalifornia

Posted 15 March 2013 - 09:08 AM

View PostThontor, on 15 March 2013 - 08:38 AM, said:

Are you saying you have never been one of those people who does early with less than 100 damage done? Because I regularly place in the top 2-4 damage done as well... Averaging at least 300... But sometimes I happen to be the guy the enemy targets and focuses fire on first...

I have said it before, but just because someone does early with little damage done doesn't make them a worse player than you, it happens to all of us... If you come back and tell me you never die early with under 100 damage done, I call BS


As much as I would like to whole-heartedly agree with you, I can't. I do top-damage a lot of the time, so it'd be easy for me to brush off the times when I'm stupid and stumble into the open. But the thing is, both in League of Legends and in MWO (I bring up LoL just because i have years of experience in that), it's those times of inattention and impatience that separate me from the the people a tier or two above me. When I die early, it is an indication of my lack of skill, because it's my own impatience that got me killed.

#94 Nightfire

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 226 posts
  • LocationAustralia

Posted 15 March 2013 - 09:12 AM

Alright, Let's do this then.

View PostMischiefSC, on 15 March 2013 - 08:25 AM, said:

Wow, ad hominem much?

No, this is a directly and attributable statement. It is completely and easily provable. Your disgust of anyone who plays primarily pre-made is documented on these forums in many places.

Quote

Look. What you're doing is effectively saying that when you lose it's because your team is terrible. Do you understand both how that sounds and statistically what that's saying? Your argument is based around the premise that the matchmaker treats you differently than everyone else. Do you understand as well how that comes across?

You're good at strawman attacks aren't you? No, I have NEVER said that when I lose it is only because of and solely attributable to a terrible team. I will clarify what I did say however, just so you can try and get your head around it.
When my team loses AND it happens that we have a good number of players on our team (3-4) that played bad enough to allow us to question if they were matched our pre-made skill level fairly then MY enjoyment suffers. This is because I feel we did not present a fair representation of our skill because several of our team were not up to the task. Our opponents usually are, but in those cases we were let down and we notice. It happens and I will admit I probably give those matches more weight than they represent but they do impact my enjoyment on the whole.

Quote

Your win/loss is still over 50% so you're climbing in Elo and will slowly but consistently be put with better and better people, so again, working as intended. Currently you've put forward a conjecture and a complaint when to all other respective points Elo is working as intended without putting forward anything but unsupported anecdotal evidence. Weight matchmaking is still too lose and PGI has said so but it's being tightened slowly. The problem with matching both weight and Elo is delays in finding matches on both ends of the spectrum and finding a good balance.

Actually, in another post I do put forward my reasoning but it is all anecdotal. What "evidence" would you like? Screenshots of matches with the losing team having 3-4 people who have a total of less than 100 damage? There are plenty of those about already; I don't need to add to those. Are you trying to tell me that watching a Raven spot, shoot at, then run directly at a Catapult A1 with 6xSRM6's is something other than my diagnosis as a bad pilot?
I'm putting forward my experiences, my observations and a theory that fits the evidence I have. I am also doing so without dismissing other peoples experiences.
What I find amusing about this theory of mine that you are so ready to dismiss as only having "unsupported anecdotal evidence" is that there are others conveying the same experience. You also dismiss my position knowing full well the only "evidence" you will accept validating my position is Elo ratings of these pilots which is not available to us. As such, in your opinion, theories that explain a portion of the user bases experiences are not permitted to be formed unless they adhere to your world view of "the pug stompers are now whining that they are losing" narrative.

Quote

Again, your teammates may not be lower in Elo than you - they may just not be good at teamwork i pugs. What Elo isn't doing is just dropping you with a bunch of incompetent boobs to try and force you to lose. The reality is that they're going to be pretty close to your own Elo level. Maybe they're all the ex-pugstompers paying their painful dues.

Firstly you have no idea of what hours I play (or the number of players available), how long we wait for matches or how much the Matchmaker will lax the parameters to beat the 2 minute time limit. If they are "pug-stompers" paying their dues, why do I not see specific players that play like this more often as they fall down the Elo scale? Your reasoning has holes.
I chose to address this point out of order because it is quite telling. I don't believe Elo isn't dropping me with "boobs" (your word) to try and force me to lose. That would require sentience and an ego, it is also telling about why you think I am contributing to this thread. What I DO think the Matchmaker is doing is on occasion it runs so short on time trying to fill the last 4 places that it grabs whoever it can with too wide a net.

Quote

What you are not doing is having a vastly different experience than everyone else in this game. Sometimes the team I drop with doesn't coordinate. Sometimes even excellent players get dropped by focused fire when the other team really brought their A game.

No, what I am experiencing is vastly different from what a majority of players on the forums report. Here on the forums there are others who have experienced play similar to mine. That both makes me not alone and corroborates a possible flaw in the Matchmaker. Additionally, this is an experience that many I talk to on the Teamspeak servers report having from time to time. Just because they don't read or post here doesn't make their experience less valid.

Quote

How many matches have you had since Elo was fully implemented? Dozens? Hundreds? If you're still gaining in Elo then you haven't settled yet and you'll continue to drop with progressively better and better people. Again. This isn't a problem, it's working as intended.

Again, you think the problem is something to do with constant matches or me not finding a settled point. Try and understand this:
  • My experience isn't your experience!
  • People other than myself have this experience!
  • This IS a problem and when this happens, it is NOT working as intended!
  • Can you not admit that there might just possibly be an issue with the Matchmaker that doesn't affect you? Or are you saying that if there are no problems on your end then everything must be working as intended?

Quote

If your solo game is bad the only question is are you winning or losing more than 50%? If you're winning more you'll continue dropping with better people. Are you saying that somehow you're consistently dropping with worse people now than when prior to Elo everyone you dropped with was completely random?

I don't play solo. I really don't like disorganised play.
I've never said or implied that I'm consistently dropping with worse people. What I am saying is that there are a low but significant number of matches that we lose because of bad matchmaking and that the numbers of matches are significant enough to frustrate myself and a number of other players such that we play far less.

#95 EgoSlayer

    Member

  • PipPipPipPipPipPipPipPip
  • Wrath
  • Wrath
  • 1,909 posts
  • Location[REDACTED]

Posted 15 March 2013 - 09:19 AM

View PostStealthsfury, on 15 March 2013 - 07:49 AM, said:

I'm getting frustrated with the fact that I have seen several matches where the bottom four in my group do under 50 dmg in a match. I'm talking zero 16 and 25. I understand there are new players that is great, but they should be no where near people who have been playing the game with a very high ELO score. I am not having fun when the system forces me to fight ridiculous odds.
Suppose you want to just mess around with a flamer hunchback. Your high ELO in that class completely screws your team. Who is that fair to?
I may have a blast screwing around but everyone on my team suffers. So I am actually pigeon holed into bringing the most optimal builds to properly represent my ELO. How is that fair to me?


Q: How does a random match maker prevent any of the above issues?
A: It can't

Follow up question 1:
Q: How long would the best player in the game last when being focus fired upon by 3-4 enemies?
A: About 5-8 seconds

Follow up question 2:
Q: If your in game lifespan is 5-8 seconds, how much damage can you deal to the enemy?
A: About one or two alpha strikes


A person with skill/experience (higher Elo) should know that flamers suck and is less likely to use them than someone who doesn't.
Higher skill players are also much more likely to focus fire and burn down enemies as a group. Being down one or two mechs can cause a group to lose quickly and make it look like a stomp because of the numeric superiority. That means a person who only got 80 points of damage isn't always a bad player or a noob.

But this assumes that the teams are balanced in capability, and this is the part that is severely broken now. People blame Elo, but it is really a separate issue because of the match maker. Teams are getting matched without even balancing by weight class in may cases due to time, or worse starting 8v5, etc. The more criteria there are to create a match, the longer it will take to fill with people that meet the criteria. And people don't want to wait several minutes for a match, when the match itself will take only a few minutes in most cases. The match maker has been skewed for time to create over team balance.

Part of this is due to the player base, a group of a few thousand doesn't have the diversity of a group of hundreds of thousands. The diversity of player's Elo coupled with the number of choices of mech compound this issue because there doesn't appear to be enough diversity in mech selection by the players to support an even distribution. It's then further complicated by 2/3/4 man groups that can group without any class or tonnage limits. These are problems that can be solved, while keeping skill ratings.


TL:DR: Elo is identifying/introducing problems with the match maker, Elo itself is not the issue.
One thing that I do think is an issue with the current Elo implementation is that the starting Elo of 1300 is too high. MWO has a somewhat steep learning curve and starting a new player with the assumption that they are average is misguided. Starting the players much lower, perhaps 900 or 1000 might help offset the "noob fillers" that some complain about.

Edited by EgoSlayer, 15 March 2013 - 09:21 AM.


#96 Aethon

    Member

  • PipPipPipPipPipPipPipPipPip
  • Mercenary
  • 2,037 posts
  • LocationSt. Louis, Niles, Kerensky Cluster

Posted 15 March 2013 - 09:20 AM

View PostMischiefSC, on 15 March 2013 - 07:54 AM, said:

I so predicted this. Even my timing was correct. About few weeks after Elo starts we get the tears over people missing pugstomping and trying to couch it as something else. That 'something else' is 'the rest of my team sucks'.

GLORIOUS.

Was looking back over stuff we discussed a month ago. Here, I'mma quote myself:

So what we have now are people used to pugstomping dropping with teams of people who, honestly? Are probably their exact same skill level but are used to being able to pug without teamwork because the other side always sucked. Now they're losing at LEAST 50% of the time and are used to winning 80%.

How about someone talking about how terrible Elo is post a screenie of their stats? Are they at 0.80 win rate? 0.30? Or are they around 1.0 still and just hating that and now looking for someone to blame?


It is not just the pugstompers. IMHO, the people sync-dropping two 4-man groups together are the pugstompers. They are bringing an entire coordinated team, intentionally, against pugs and mixed-groups. The typical excuse I hear is that they cannot find 8-man matches, which is no excuse IMHO.

A lot of the better players here only run 2-3 man groups, because we end up with 6 people online instead of a full 8 (which is when we usually do 8-man drops). You also do not take into account that we have to level crappy mechs like the Raven 2X and 4X just like everyone else, and we always do that in our pug drops...because doing that in 8's is suicide.

When the Bears do 4-mans, we take fun builds, not one of the 5 hardcore builds you are forced to use in 8-man groups. We take stuff like Dragons, Cicadas, hero mechs, test builds, crappy mechs we need to level, etc., because it is fun. Unfortunately, ELO has changed that. Over 90% of the time 4 of us get together on a team, the other 4 team members are bads. This turns the game into 4v8, because the bads never do anything useful before dying. This is a repeatable problem, that happens match after match after match.

Frankly, it is not fair to the other 4 players on our team, either. They are being dragged up into a match where they do not belong, and expected to learn somehow. What was that people were saying about ELO being good for the new player experience? lol Sure, ok. It looked REAL fun for the people stuck on my team...for the 90 seconds their match lasted.

Oh, and before you claim that I am one of the 'pugstompers' to which you were referring...I was one of the people who suggested giving pugs their own queue, where groups could not enter...and several of the pugs shot back that they did NOT want this. I was one of the people requesting that new players only be matched against each other for their first 25 matches or so, and I also requested private matchmaking where they can practice with each other, try new stuff, and the veteran players can actually teach them new things.

Edited by Aethon, 15 March 2013 - 09:22 AM.


#97 MischiefSC

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • The Benefactor
  • The Benefactor
  • 16,697 posts

Posted 15 March 2013 - 09:25 AM

View PostYokaiko, on 15 March 2013 - 08:55 AM, said:



Then that will be the end of me playing, because its no longer fun.




Why where you out in the open?

That is mechwarrior 101.....don't be very far from cover.


I don't believe you.

For one thing it's still better than pre-Elo. The only difference now is that dropping with a 4-man team doesn't give you an 80% win/loss rate no matter what your other 4 teammates are like. You have to coordinate, especially when you pug.

What Elo is doing isn't giving you worse teammates - it's giving you better enemies. That's not going to change. I find it hard to believe that you're going to quit because of that. Sorry man, not calling you out, but really? You're upset because you feel like your teammates are sandbagging you. I had almost 20 straight losses day before yesterday because I couldn't catch a team that would coordinate. Some of them were premades even, all the same wacky black and green paint jobs. Refused to use chat, a couple people would and everyone else would just do something else, abandon their LRM support, leave the Atlas brawler behind and rush in medium/heavies, the list was long and painful.

Yesterday and today have been better. Elo, like any matchmaking, works on the aggregate. You had a few bad games. It happens. Keep going. Keep your win/loss over 50% and it will get better.

Where you start to hit suck is when you run into the ex-pugstomping teams who are trapped under the heavily competitive teams.

We discussed this a while ago but it's worth bringing up again:

Elo approximate scale -

Tier 1 - Golden Ninja Elite.

Best teams in the game. Probably loving Elo or at least not minding it but probably playing it less because every game is like competitive 8-mans. They will almost perpetually have a greater than 50% win/loss rate because statistically they are too few to only drop with comparable skilled players so lower Elo folks get brought in. Winning against a lower ranked player doesn't raise your Elo much if any so they win a lot of games for only a tiny increase.

Tier 2 - Ambrosia.

Ambrosia is the food of the gods. The problem is that you ARE the Ambrosia at Tier 2. You are who gets beaten, often, by the Tier 1 folks when you pug. You'll get pulled in to fill out their 4man drops when there are not enough Tier 1 folks around. If you're a competitive team player yourself this isn't bad - you'll continue to learn and improve until you as well join the Golden Ninja Elite. This is the hard grind zone - if you're a pug here any team that gets you should be happy indeed, you probably know all the tricks and have gotten by on skill alone. If you're a team member but pugging, this can be a painful place to play. You'll be going solo in games involving competent competitive teams and very skilled players. If you don't work hard to coordinate you're going to get rolled a lot.

Tier 2 - sub-basement, The Land Of Spoiled Milk and Bitter Honey.

These are the guys who use to pugstomp and sync drop to pad their stats and are now in a hell they can not escape. They were never very good but they coordinated cheese builds with buddies and pugstomped like mad, relying on the advantages Teamspeak gave them over disorganized pugs to reap many victories with little effort. Suddenly Elo happened and the train stopped, kicking them out at a station full of cold-killin' ******** and they are getting effectively crushed all the time.

These are the guys who are getting you killed. They ended up in Tier 2 not by dint of personal skill but via exploiting tricks against less skilled opponents. Especially when they pug (which is getting more common as the game is probably not as fun for them and dropping with their teams to get ROFLSTOMPED by Tier 1 folks isn't what they played for) they're going to play like idiots but in cheese builds, still hoping for some easy kills against someone having an off day.

Statistically they are what I suspect you're seeing. It's been a month and the people who used to pugstomp for 80% wins have been getting crushed for a while and are splitting up from their old 4-man drops. Now they're pugging and they suck at it. They don't communicate or coordinate, they assume their team are all idiots, when they do 4-man they try to use their pug teammates as fodder - which doesn't work because pugs at this level are, as you sound like, experienced players who have no intention of being fodder. So they just suck.

Tier 3 - Glorious Puggie Heaven

This is the best place to pug. The only premade teams here are guys out playing for fun, they never focused on competitive play. Skilled pugs play here and generally like to coordinate. Still get pulled into T2 or even T1 games as padding but do alright - if people coordinate with you.

Tier 4 - Noobzville

New player start here. This is neutral Elo and it's a pretty wide band. Almost everyone here improves to move up to T3 unless they suck enough to end up in .....


Tier 5 - Rambozone

Teamwork? Pfft. That's for losers. If I lose it's my teams fault for not doing better. I'm amazing, I don't need to use chat. I AM AN ARMY UNTO MYSELF. I'll just wander out into the lake alone and snipe at the enemy. THEY WILL ALL FEAR MY PPCS!

Tier 6 - Steering Wheel Underhive

Don't open the curtain. Don't do it. You don't want to see what happens here. That which has been seen can not be unseen.


I suspect you're around Tier 2 and running into a lot of the folks from the land of Spoiled Milk and Bitter Honey. They'll sort out soon enough. You just need to win enough to rise above them. They will continue to suck and will fall down to Tier 5 where they belong. They just have to lose a lot of games.

How many games have you played since February? Hundreds? 50? It matters - a lot.

#98 FerretGR

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,445 posts

Posted 15 March 2013 - 09:27 AM

View PostAethon, on 15 March 2013 - 09:20 AM, said:

Over 90% of the time 4 of us get together on a team, the other 4 team members are bads.


Over 90%? You're being precise, I can only assume you've collected some data? Care to share?

If not, it's "over 90% of the games I care to remember" which strongly indicates confirmation bias to me.


View PostMischiefSC, on 15 March 2013 - 09:20 AM, said:

post


I'd be surprised if your analysis of the situation isn't at least somewhat correct, Mischief!

Edited by FerretGR, 15 March 2013 - 09:33 AM.


#99 Galenit

    Member

  • PipPipPipPipPipPipPipPip
  • 1,198 posts
  • LocationGermany

Posted 15 March 2013 - 09:29 AM

200 Matches:
Assault: 97 wins, 98 losses, 197 matches
Conquest: 1 wins, 1 losses, 3 matches

For me, elo is working.
Dont know where the third match under conquest is gone ....

But the maps are not as random as i wish:
Forest Colony: 21
Foerst Colony Snow: 24
River City: 38
River City Night: 21
Frozen City: 23
Frozen City Night: 31
Alpine Peaks: 27
Caustic Valley: 15

Maybe it evens out in the greater numbers ...

#100 Nightfire

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 226 posts
  • LocationAustralia

Posted 15 March 2013 - 09:31 AM

View PostMischiefSC, on 15 March 2013 - 08:50 AM, said:

Thontor brings up a great point - you've never been killed with less than several hundred damage? Ever? You're never not the top of the list?

Have I been killed early? Most definitely yes!
Have I been killed with less than 100 damage when I've been killed early? Not that I can recall. I'll admit it is possible but I do look for this sort of thing to evaluate and reflect on my performance.

Quote

For the arguments that you and Nightfire are making to hold statistical significance you guys would have to NEVER have a bad game. Ever. You play perfect every game and are always the best performer every time you drop.

Do you see how that comes across?

I see how you come across when you misrepresent my argument like that! Furthermore, I've taken the time to actually ASK people about their experiences before forming my theory or coming to the forums. You are misrepresenting my position, making generalised statements that only hold true for your experience and implying those that hold my position are simply complaining.
May I ask, how did you respond when you complained about "pug stomping pre-mades" and you were similarly dismissed with something like, "Get on Teamspeak and get a team"? Or does that not matter because your experience is valid and ours is not?

Quote

What is more likely is that you're all speaking to a form of cognitive bias like availability heuristic. You're focusing on the times other people have done poorly and you have done well and it's skewing your statistical measurement. If your win/loss is over 50% then actually most the time your team is doing great. The other team is probably looking a lot like your team does when you lose - your team gels, focuses well and coordinates and the other team dies before they even have a chance to do much.

This speaks the most that you simply have no clue as to what we are speaking about. You haven't experienced it and it seems to me that as such, it simply cannot be valid. Thus, "whining, arrogant pug-stompers".

Quote

High Elo matches are going to involve a lot of very good players. It's not going to pull in noobs. The problem with high Elo players though is that when they're good and they coordinate they're freaking amazing. When they're good and they don't, then they're just good. One side coordinates well, otherside doesn't, you're dead before you even had a chance.

And you know this how exactly? Again, you have no idea what time I generally play, how many players are available, how long we generally wait for a match or how lax the Matchmaker can get with the rules.
What you are speaking from is your experience and dismissing anything that doesn't match up to it.

Quote

Elo. Working as intended. The statistical probability of you dropping with people who suck is so low as to be almost insignificant. More likely you're dropping at times when game populations are low and you're seeing a wider swing though and this makes the relative win/loss swing wider.
You are however have the exact same game experience as everyone else.

Not working as intended in all instances.
Your absolutist statement is already falling apart. I reject your statement that I am the only one with this experience. Now Yokaiko expresses a similar experience so you are forced to revise your position to just the two of us and no one else.
The cracks in your complete denial position are getting larger.





7 user(s) are reading this topic

0 members, 7 guests, 0 anonymous users