540 replies to this topic

[YueFei]

wanderer, on 01 February 2014 - 02:15 PM, said:

Right. So when the time came in a real match, were they slapping your lower-skilled self in against the best, or were they putting your state champs against whoever was best on the other side? Being that actual competitions are where it matters.

Training is one thing. "Losing" don't matter, learning does. MWO, a loss matters. It's a slower road to any kind of progress- pilot skills, new equipment, whatever. Without even the taste of being able to perform consistently and succeed, the general result is "Hi, I'm a scrub and I'm here to be used for target practice".

You, and I know what it's like to succeed. New players are learning what it means to be beaten repeatedly instead, courtesy of the "state champs" of MWO.

A robust single player would go a long way towards teaching newer players and letting them practice. Just like most fighting games let you learn the ropes against the AI.

There are 3 "zones" for doing things. I don't remember the exact terms, but the first zone is too easy and you learn nothing. The "stress zone" is where you are forced to do something you are not used to, to learn, to grow, and to improve, and this is where the most learning happens. Then there is the "panic zone" where things are so far over your head that you're learning nothing.

Dropping new players against the top veteran players puts those new players in that last zone. Even if they have the right attitude, are patient, and don't mind losing... the problem is that things are so overwhelming that they're not learning anything. I have a friend who we were playing with, a brand new player, and despite us trying to guide him and tutor him, he wasn't learning anything. He has been very resilient and persistent, has a good attitude, and tries hard, but so often we'd watch him die and he would not know why he died or what mistake he had made. Eventually he stopped playing with us and tried to solo drop, under the theory that the matchmaker would pit him against other novice players. But that hasn't helped either. Even when he wins a game, so often it's because the matchmaker put aces on his team and they cleaned up the enemy before he even realized what was going on.

A robust single player campaign, along with strong practice "challenges" to help hone specific skills, is really sorely needed. They really ought to invest more resources into this, even if only for their own financial gain! The new players it would help to retain would make it more than worth their while.

[Nick Makiaveli]

A banana in the tailpipe, on 01 February 2014 - 12:59 PM, said:

LOL first match today, a total stomp. Not even a single assist from my team. (those jerks wouldn't let me duel a raven in peace and were only out for blood) I thanked them for being the reason this game is tanking fast.

You are upset they wouldn't honor your duel and instead sought to win the game and you say they are the reason the game has problems?

Didn't occur to you that you are being an {Dezgra} and trying to solo play a team based game? Of course it doesn't go as well.


Starting to wonder if you are just trolling, or if you have helicopter parents and thus have no clue what the real world looks like.

[Gruinhardt]

A banana in the tailpipe, on 31 January 2014 - 03:59 PM, said:

I backed up my research and data with facts. Those screens and my personal experience over the past 48 hours conclude the matchmaker is borked and player's claims of constant "stomps" are true. PGI can ignore it and bleed players or offer incentives to bring fresh blood into the game so that there is a greater pool for the matchmaker to draw from. PGI is either too cheap or stupid to do this.

Have a nice day.

Would it make you feel better if I posted a bunch of screenshots of close games. I could post 9 screenshots of close games, and 1 stomp. My facts would show stomps only happen 10% of the time.

[Sandpit]

I'm just going to say this

Anyone who thinks the OP's "facts and data" are anythign but anecdotal you really should take a stats class.

40 matches out of the THOUSANDS played is not a good sample size. Nor is it a good indicator. 1 player's experience is not "statistical proof" of anything. If you REALLY want statistical data? Get about 500 players together(that's 500 pugs and 500 premades), take the data collected over about 200-300 matches (for each of them) and THEN you have a small sample size compared to all of the games played since CB.

THEN you have somethign to start basing opinions on. THEN you have a small but decent sample size. THEN you can start to see trends. 40-50 matches by one player simply is not enough. If you believe otherwise you're either just too stubborn to listen, ignorant (note I said ignorant, not stupid or dumb) of how statistics work and how data collection is actually performed.

Then you can run that data through SAS, NCSS, Excel (if you can't get the others), etc. and see what the data actually represents. You can argue all you want but THAT is a factual way of how you'd begin to collect factual data with a large enough sample size to actually denote a few trends. Anything else is simply too small and too anecdotal to call "fact"

If there are 20,000 players in this game (active meaning they drop at least once a week) for example, you'd want at LEAST 2000 players.

Then if you're taking random samples you're still not going to get very accurate data. You won't get it because that's the worst way to sample. IF you REALLY want data then you collect data from players who don't know they're being monitored so they don't adjust the way they play to skew results.

Now, with all of that said, I wonder who could possibly have that kind of data...
I wonder who could possibly collect and track data across not one, but every single player in the game. I wonder who would benefit the most from making decisions regarding queues. Would the PUGs calling for them benefit the most? Would those in premades saying it can't be done benefit the most? Or, would the company banking the biggest project of their collective careers benefit the most making adjustments based on data collected from the ENTIRE game (as opposed to "proving" something from a single player's experience across 40 matches) I do so wonder..........

The thing is, most people who drop in premades that I've seen have never EVER said, "We dont' "WANT" separate queues." (in direct contrast to those who want to jump up and down and claim premades demand anything), they've said, "It just wouldn't work because of the launch times based on what we've seen" (Hmmmmm, let's see if you're paying attention. Would this be anecdotal just like the OP's "evidence") yes, yes it is anecdotal JUST LIKE THE OP

PGI has made several adjustments since CB regarding the whole premade, grouping, and pug thing. They've nerfed groups, they've implemented ELO, they've reduced how you can even group up. So tell me again where the "evil PGI" machine has left the pugs out to dry and hasn't made several adjustments to placate pugs that complain about this over the years?

So with that said, why would anyone who is being unbiased and objective think that PGI is "forcing pugs to get stomped"?

Nick Makiaveli, on 01 February 2014 - 03:08 PM, said:

Starting to wonder if you are just trolling

That was apparent the first time he said "Look, 4 screenshots over 40+ matches! This proves all games end in stomps"

[Sandpit]

YueFei, on 01 February 2014 - 03:03 PM, said:

Dropping new players against the top veteran players puts those new players in that last zone.

Not to discount anything else you said but this was the most important part in my opinion. I've been saying this since CB. New players are the ones that should get the own queue. They should only be dropping against other "cadets" and be able to get comfortable with the game before they drop into a harder setting. I think this would go a loooooooooooong way to mitigating stomps and prevent new players from getting a bad taste in their mouths along with helping retention.

[Sandpit]

A banana in the tailpipe, on 01 February 2014 - 11:22 AM, said:

The rest of you who want to thump your chests and scream I'm wrong are just jelly I thought of a cooler call sign than you did.

This is the problem. I've never said you're wrong. I've said your data collection is wrong and therefore the assertions based on that data is incorrect. I've posted why. I've posted exactly why a single player (regardless of who it comes from, you or otherwise) dropping in 40+ matches and psoting screenshots from 4 games that ended in stomps is not an indicator of every game ending in s stomp, or a majority, or a minority, or any other such demographic. Your evidence is simply anecdotal. The fact that you can't accept or understand that just shows you don't understand how statistics and data gathering actually work when you're talking about something as large as this game is. It's not a matter of "you being wrong", it's a matter of that data collection simply isn't enough and isn't going to be anything more than your personal experience.

It doesn't take into account the thousands of other players and their experiences. It doesn't take data from the thousands (not 40+) matches being played. It doesn't take into account the difference in those stats regarding the thousands of ELOs, weapons, mechs, premades, pugs, etc. that play this game.

If you REALLY want your data to be taken seriously I've laid out steps on how to do that. Instead your'e simply saying "I'm right, I have proof, anyone who disagrees is wrong. Anyone who doesn't accept my 4 screenshots as proof that every game ends in stomps is a white knight and ignoring "facts"

*shrugs*
Keep doing what you're doing though. You seem to be having fun and every match you drop is more data for PGI to look at

[Craig Steele]

Sandpit, on 01 February 2014 - 06:52 PM, said:

This is the problem. I've never said you're wrong. I've said your data collection is wrong and therefore the assertions based on that data is incorrect. I've posted why. I've posted exactly why a single player (regardless of who it comes from, you or otherwise) dropping in 40+ matches and psoting screenshots from 4 games that ended in stomps is not an indicator of every game ending in s stomp, or a majority, or a minority, or any other such demographic. Your evidence is simply anecdotal. The fact that you can't accept or understand that just shows you don't understand how statistics and data gathering actually work when you're talking about something as large as this game is. It's not a matter of "you being wrong", it's a matter of that data collection simply isn't enough and isn't going to be anything more than your personal experience.

It doesn't take into account the thousands of other players and their experiences. It doesn't take data from the thousands (not 40+) matches being played. It doesn't take into account the difference in those stats regarding the thousands of ELOs, weapons, mechs, premades, pugs, etc. that play this game.

If you REALLY want your data to be taken seriously I've laid out steps on how to do that. Instead your'e simply saying "I'm right, I have proof, anyone who disagrees is wrong. Anyone who doesn't accept my 4 screenshots as proof that every game ends in stomps is a white knight and ignoring "facts"

*shrugs*
Keep doing what you're doing though. You seem to be having fun and every match you drop is more data for PGI to look at

But the thing is that this is ONE of Hundreds of different people, maybe thousands over the last year (if others are accurate) which is a statistic PGI should be concerned about.

Argue the pro's and cons of anecdotal evidence till the cows come home. Argue it's biased. Argue away.

It's not going to change that hundreds of DIFFERENT players are saying they do not enjoy stomps, and statiscally another 100 more players for each of them is leaving the game without saying anything.

The argument that compainers represent a small portion of the population may well be correct, but those putting it forward have no evidence that is the case. It might be that 90% of the population are activily posting their stomps. The argument on is no stronger than anyone saying "all games are stomps". Put up a screen shot showing its not and all the arguments are just as valid reversed.

But what seems clear is that there are more people posting a concern than saying "hey this is WAI"

The only real people who can ensure clarity around the question are staying silent except to say, we'll keep the data secret but trust us, it's WAI. I don't get any comfort from the PGI communication.

[Sandpit]

Craig Steele, on 01 February 2014 - 09:05 PM, said:

But the thing is that this is ONE of Hundreds of different people, maybe thousands over the last year (if others are accurate) which is a statistic PGI should be concerned about.

Argue the pro's and cons of anecdotal evidence till the cows come home. Argue it's biased. Argue away.

It's not going to change that hundreds of DIFFERENT players are saying they do not enjoy stomps, and statiscally another 100 more players for each of them is leaving the game without saying anything.

The argument that compainers represent a small portion of the population may well be correct, but those putting it forward have no evidence that is the case. It might be that 90% of the population are activily posting their stomps. The argument on is no stronger than anyone saying "all games are stomps". Put up a screen shot showing its not and all the arguments are just as valid reversed.

But what seems clear is that there are more people posting a concern than saying "hey this is WAI"

The only real people who can ensure clarity around the question are staying silent except to say, we'll keep the data secret but trust us, it's WAI. I don't get any comfort from the PGI communication.

No, no it's not. That's the entire point of my posts. Even if you take the face value of the statistical data collected it showed 4 stomps out of 40-50 games. That's simply not adding up to "every game" or even "half, the majority, a lot" end in stomps. That's roughly 10 or less %. That's my point. Everyone wants to jump up and down about "look here's 4 screenshots, this proves it" when it just doesn't. What it does prove is that player in an alt account, running nothing but trials experienced 4 stomps in 40-50 matches over the course of 48 hours.
4 matches out of 40-50 isn't even what I'd call a lot. That's a minority actually

Even if it is hundreds of different players experiencing stomps abotu 10% of the time? That's simply not a majority, nor is it any kind of majority much less "every game"

There is no "arguing" anecdotal evidence. That's what it is. Regardless of your opinion about it, 40-50 matches by one player ending in 4 stomps is NOT a significant amount of data compared to the tens of thousands of matches played. it's just not. That's not how statistics work regardless of individual opinion. It's not an argument. it's pointing out that 40-50 matches by one player over 48 hours is simply anecdotal.

Quote

The argument that compainers represent a small portion of the population may well be correct, but those putting it forward have no evidence that is the case. It might be that 90% of the population are activily posting their stomps. The argument on is no stronger than anyone saying "all games are stomps". Put up a screen shot showing its not and all the arguments are just as valid reversed.

Exactly. Except the ones on the other side of the argument aren't calling for sweeping and fundamental changes to the game. Those complaining are, based purely on anecdotal evidence. That's the difference.

Does the MM need some adjustments? I'm sure it does. I'm sure it can be tweaked and adjusted here and there. I also know that PGI is perfectly aware of this and researching the data to make those adjustments. I know this because they flat out said they have been.
Things like "possible elo reset"
"Adjustments to the MM system and elo this weekend"
"Tonnage restrictions"
and a myriad of other statements over the years show me that they are taking steps to get closer matches more often. The fact that an person's opinion on whether or not to believe that is irrelevant because if you do believe it, you really have no reason to post "every game ends in a stomp and PGI isn't listening and won't fix it" and if you don't believe it, then a post of "every match ends in a stomp and PGI isn't listening and won't fix it" is pointless any way because they aren't listening to you in the first place.

Everyone is entitled to their own opinion. If you're going to post 4 screenshots and say that's conclusive evidence that every game ends in a stomp then you might expect someone who understands statistics and data gathering to point out why that ISN'T in fact conclusive evidence and doesn't show that every game ends in a stomp. It also gives a really bad perception to new players that, according his own data, 10% of the games played (or less since it was 4 screenshots and we don't know the exact number of matches played that DIDN'T end with a stomp) indicates that every game ends in a stomp.

[wanderer]

Sandpit, on 01 February 2014 - 09:29 PM, said:

Everyone is entitled to their own opinion. If you're going to post 4 screenshots and say that's conclusive evidence that every game ends in a stomp then you might expect someone who understands statistics and data gathering to point out why that ISN'T in fact conclusive evidence and doesn't show that every game ends in a stomp. It also gives a really bad perception to new players that, according his own data, 10% of the games played (or less since it was 4 screenshots and we don't know the exact number of matches played that DIDN'T end with a stomp) indicates that every game ends in a stomp.

And that bad impression will continue until the matchmaker is fixed, Sandpit. No matter what you say, what you do, how you chop logic- perception is all in the gaming industry. And the "stomps keep coming" threads will continue to pop until PGI drowns in bad publicity. It's not 4. It's a constant stream of these threads, and they -will not stop until there is a change-. Promises don't do, we're "live" now, remember? Can't call it a beta. Live. Like this.
Please, keep saying it isn't so. A "sand pit" is often used to dispose of dangerous unexploded ordinance, after all.

[Craig Steele]

Sandpit, on 01 February 2014 - 09:29 PM, said:

No, no it's not. That's the entire point of my posts. Even if you take the face value of the statistical data collected it showed 4 stomps out of 40-50 games. That's simply not adding up to "every game" or even "half, the majority, a lot" end in stomps. That's roughly 10 or less %. That's my point. Everyone wants to jump up and down about "look here's 4 screenshots, this proves it" when it just doesn't. What it does prove is that player in an alt account, running nothing but trials experienced 4 stomps in 40-50 matches over the course of 48 hours.
4 matches out of 40-50 isn't even what I'd call a lot. That's a minority actually

Even if it is hundreds of different players experiencing stomps abotu 10% of the time? That's simply not a majority, nor is it any kind of majority much less "every game"

There is no "arguing" anecdotal evidence. That's what it is. Regardless of your opinion about it, 40-50 matches by one player ending in 4 stomps is NOT a significant amount of data compared to the tens of thousands of matches played. it's just not. That's not how statistics work regardless of individual opinion. It's not an argument. it's pointing out that 40-50 matches by one player over 48 hours is simply anecdotal.


Exactly. Except the ones on the other side of the argument aren't calling for sweeping and fundamental changes to the game. Those complaining are, based purely on anecdotal evidence. That's the difference.

Does the MM need some adjustments? I'm sure it does. I'm sure it can be tweaked and adjusted here and there. I also know that PGI is perfectly aware of this and researching the data to make those adjustments. I know this because they flat out said they have been.
Things like "possible elo reset"
"Adjustments to the MM system and elo this weekend"
"Tonnage restrictions"
and a myriad of other statements over the years show me that they are taking steps to get closer matches more often. The fact that an person's opinion on whether or not to believe that is irrelevant because if you do believe it, you really have no reason to post "every game ends in a stomp and PGI isn't listening and won't fix it" and if you don't believe it, then a post of "every match ends in a stomp and PGI isn't listening and won't fix it" is pointless any way because they aren't listening to you in the first place.

Everyone is entitled to their own opinion. If you're going to post 4 screenshots and say that's conclusive evidence that every game ends in a stomp then you might expect someone who understands statistics and data gathering to point out why that ISN'T in fact conclusive evidence and doesn't show that every game ends in a stomp. It also gives a really bad perception to new players that, according his own data, 10% of the games played (or less since it was 4 screenshots and we don't know the exact number of matches played that DIDN'T end with a stomp) indicates that every game ends in a stomp.

We agree to disagree then.

You seem to be arguing that if one player posts 4 screen shots and says "hey my experieince in this game is poor and here's some examples of what I am talking about" then we should be aware that he is only ONE player (and others have argued he is biased etc etc) I don't dispute that.

I am saying that he is not the only player, and that hundreds (some would say thousands) of people are saying the same thing. I am saying that for every player saying this there is a big number how say nothing. I am saying that anyone saying the game is WAI is using the same argument as the poster here except in reverse. The argument is no stronger because it relies upon exactly the same evidence, that players experience.

PGI say "it's all top secret but trust us, it's WAI"

I say to PGI, give us something more meaningful so we can have confidence it is WAI cause lots of people are still expressing the same thing.

If you're saying the post title and actual fact are different (ie, you're concern is the title) then yeah. Even if all his games are stomps that not every game ever played obviously. But I suspect the title is more about sentiment.

Edited by Craig Steele, 01 February 2014 - 09:44 PM.

[Sandpit]

wanderer, on 01 February 2014 - 09:36 PM, said:

And that bad impression will continue until the matchmaker is fixed, Sandpit. No matter what you say, what you do, how you chop logic- perception is all in the gaming industry. And the "stomps keep coming" threads will continue to pop until PGI drowns in bad publicity. It's not 4. It's a constant stream of these threads, and they -will not stop until there is a change-. Promises don't do, we're "live" now, remember? Can't call it a beta. Live. Like this.
Please, keep saying it isn't so. A "sand pit" is often used to dispose of dangerous unexploded ordinance, after all.

So 10% of games ending in a stomp is now indicative that every game ends in a stomp?

It's a constant stream of these threads by the same people on both sides with a couple of new players joining each side every "new" thread.
It also completely dismisses everythign said about PGI already taking steps to fix that problem for the 10%
It's isn't "chopping logic" it's cold hard factual math. it's how statistics work. It's not opinionated or subject to perception. The screenshots and evidenced provided does NOT indicate every game ends in a stomp. period. It doesn't even indicate half or one quarter of games end in a stomp.

that's just like me saying "I went to McDonald's 50 times over the past week and 4 times my fries were burnt. Every time I order fries from McDonald's they're burnt" even if a thousand other people agree and experience the exact same numbers (which is HIGHLY unlikely) that's STILL not indicative that every order of fries ordered is burnt. Nor does it indicate a majority are. It might ***** my personal opinion (if 10% of the time my fries being burnt bothers me) but that in no way even remotely indicates that the majority of McDonald's customers suffer the same thing.

[Sandpit]

Craig Steele, on 01 February 2014 - 09:41 PM, said:

We agree to disagree then.

You seem to be arguing that if one player posts 4 screen shots and says "hey my experieince in this game is poor and here's some examples of what I am talking about" then we should be aware that he is only ONE player (and others have argued he is biased etc etc) I don't dispute that.

I am saying that he is not the only player, and that hundreds (some would say thousands) of people are saying the same thing. I am saying that for every player saying this there is a big number how say nothing. I am saying that anyone saying the game is WAI is using the same argument as the poster here except in reverse. The argument is no stronger because it relies upon exactly the same evidence, that players experience.

PGI say "it's all top secret but trust us, it's WAI"

I say to PGI, give us something more meaningful so we can have confidence it is WAI cause lots of people are still expressing the same thing.

If you're saying the post title and actual fact are different (ie, you're concern is the title) then yeah. Even if all his games are stomps that not every game ever played obviously. But I suspect the title is more about sentiment.

What I'm saying is that you can't base balance on such a small sample size. you can't base fundamental changes for the entire player population based on anecdotal evidence. That's my point. It's not personal to anyone, it's a matter of me posting why 4 screenshots over 40+ games isn't conclusive evidence of anything.

Not to mention, since we're talking statistics, if you want your data to be taken seriously you need to provide MUCH more information over and above "I've played over 40 games (see how there's no definitive number there which makes basing ANY kind of stistical determination impossible?) in 48 hours and here's 4 screenshots that show stomps"

Let's break that down
40+ games: Well we don't even know the total games played so we have no idea what percentage of games were stomps
How many games WEREN'T stomps? We don't know because again, data collection was poorly done so we're working on incomplete data

The fact that in one of those games an entire lance did less than 100 damage is interesting. We can't put it into context because we don't know how that happened. We don't know if they were poor noobs dropped in against vets and were just outclassed, we don't know if they were rambos that just ran in and died to farm some c-bills, we don't know if they just had a bad round.

We don't know the ELOs of the players involved so the claims that "I was dropped against high ELO players" is simply untrue because that factor is completely unknown. (And no, the whole "I know someone's ELO because they're "good" doesn't hold water because we have NO idea of our ELOs due to k factors, wins against lower players, losses against higher players, etc.)

Simply put, I was pointing out that no matter how hard some may want it to be, it's simply is not enough evidence to prove anything. it's not a matter of my opinion on it, it's just a matter of 4 screenshots and some vague descriptions of the number of matches played, etc. isn't conclusive evidence of anything

I'm not saying stomps don't occur and that they're not an issue. What I AM saying is that it's not always the MM that's responsible for them. Just as many times as I've seen MM bork me with weight differences or just simply dropping me into a match where I really felt like the other team was just well above my skill level, I've also seen 6-8 players run around in a circle like benny hill chasing a spider while LRMs rain down on everyone, or 2-3 rambos run out and get killed because they're just playing REALLY badly, and numerous players just discoing leaving my team 1-2 players down from the start (sometimes disco because they don't like the map, others because of tech problems) oh AND the number of times I've watched 2-3 AFK teammates drop into a match and do absolutely nothing but take up a spot on the team.

Point being it's not always PGI, MM, or premades. Sometimes (i'd hazard a guess of about 50% of the time but as with anything it's anecdotal because I don't have actual stat tracking on it) it's the players.

[Deathlike]

Same size of just the matches I had today: Pretty much stomps, either winning or losing.

"Fun"

My definition of stomp is pretty much 12-6 or worse (like 12-0), but the games don't feel challenging... you either await your fate or pound away. I guess that is my ELO level doing the work.

[Sephlock]

Deathlike, on 01 February 2014 - 10:27 PM, said:

Same size of just the matches I had today: Pretty much stomps, either winning or losing.

"Fun"

My definition of stomp is pretty much 12-6 or worse (like 12-0), but the games don't feel challenging... you either await your fate or pound away. I guess that is my ELO level doing the work.

12-6 is a stomp? That doesn't seem THAT bad.

This one was 12-5, was THIS a stomp?

[Craig Steele]

@ Sanpit (I didn't quote for length )

Firstly I'm not sure if your "personal" comment is hedging that you thought I was upset, I'm not, robust discussion here, all good:)

I can agree with you that any one person saying "All games are stomps" is hardly a cause for sweeping changes to MM.

But I submit that anyone saying WAI so "deal with it" has an equally invalid argument, if only because they can only speak from their experience.

You allude in one post that the people who are best equipped to monitor and assess this are PGI, correct. But I don't see them providing any reasonable reassurances as to its quality. The most glaring one to me was to tell us that even though the maximum elo difference in a MM game was 750, they were adjusting the maximum search significantly higher. Even my simple mind can work out that there is no reason to have a threshold double+ the maximum, so why do it?

I'm just not convinced that the number of people posting similar sentiments should be discounted. My mind screams that they are probably a minority but there is no evidence of that, the "I'm being stomped" view might be 50% of players atm. We just don't know. One man's stomp is another man's challenge too, so the definition is something that still needs to be factored.

Personally I don't focus on the actual score. It's more about the feel. If we are down 8 - 0 in 2 minutes and it ends up being 12 - 5 cause the last lance fought outside the skill set, I don't feel any better about my game experience. It was a stomp to me whether I am in the last lance, already dead or on the winning side.

If I win or lose 12 - 2 at 12 minutes but it was a tough battle of maneuver or a drawn out sniper fest, I got my money's worth, I can say that was a good game and one side didn't take their chances.

Yes there are reasons for stomps, but if MM is doing a good job then that rambo lance will be on my side as often as I pick them off. Those lemmings following the guy who says "group up" will sometimes be under my sights one by one just as much as lined up behind me. Why, because I will be in my skill set.

[Wispsy]

Sandpit, on 01 February 2014 - 09:58 PM, said:

We don't know the ELOs of the players involved so the claims that "I was dropped against high ELO players" is simply untrue because that factor is completely unknown. (And no, the whole "I know someone's ELO because they're "good" doesn't hold water because we have NO idea of our ELOs due to k factors, wins against lower players, losses against higher players, etc.)

Players with 1000s of games since the introduction of it and win/loss ratio over 3 have high Elo...you can be sure of this.

[Sandpit]

Craig Steele, on 01 February 2014 - 11:29 PM, said:

snip

Eh, it's mostly a matter of opinion when it comes to PGI more than anything else. I choose to believe that PGI does what they feel is best when it comes to the data they have. I also firmly believe that most complaints about stomps come from players who get stuck in a back bracket for them and run into a string of stomps at one time which pulls them into another bracket, or they take a break and it puts them into a different section, whatever, and then they go back to what their individual skill level dictates.
Then if they're not a great or maybe even slightly below average their "normal" win record isn't positive or right at even and so they lose a fair amount of games until they hit another string and the process repeats.

There's also so many other factors that go into it. Time of play (prime times? weekends? euro times? a mix?) or whether you're a pug or in a group, or what weight class you're dropping in. Then there's also the question of what mode you're in. There's just WAY too many variables that anecdotal evidence like earlier is enough to determine anything.

Then it comes back to opinion of PGI and how they're handling things.

Again, I personally believe, given the history I've seen, is that they truly do cater a bit more to the pugs in this instance. They have repeatedly put things in place that prevented players who like to group up from playing (IE nerfing the group sizes) and implementing things like elo.
If they didn't care about pugs or see them as important they wouldn't have done any of that. But, again, that's personal opinion on PGI.

Wispsy, on 01 February 2014 - 11:41 PM, said:

Players with 1000s of games since the introduction of it and win/loss ratio over 3 have high Elo...you can be sure of this.

No, no you can't. You can't because you don't know your opponents' elo scores. That means you could have won hundreds of matches with absolutely no movement on your elo. You also could have hundreds of losses that were lost to lower elo scores which took a major chunk from your elo.
You can't. It's impossible.

[Wispsy]

Sandpit, on 01 February 2014 - 11:44 PM, said:

No, no you can't. You can't because you don't know your opponents' elo scores. That means you could have won hundreds of matches with absolutely no movement on your elo. You also could have hundreds of losses that were lost to lower elo scores which took a major chunk from your elo.
You can't. It's impossible.

According to PGI average difference between teams is 75-125. Even if the vast majority of games are so wild the change is only one point of change...does not matter...I mean we are talking at least 3000-5000 games here, if not more.


Edit: Also it has been noted on streams once or twice certain people hit cap. By IGP/PGI employees.

Edited by Wispsy, 01 February 2014 - 11:51 PM.

[Craig Steele]

They definitely cater to the PUG imo, and so they should.

PUG to me = casual gamer looking for 2 - 6 hours of entertainment twice a week, that's the market that drops a few hundred bucks to "catch up" what the grinders / Pro's have got from "in game"

I can't fault PGI for catering to that demographic commercially.

[wwiiogre]

Sandpit, I have dropped thousands of matches since elo started and matchmaker stopped using tonnage as a basis and instead went to win/loss. Since that is really what elo is, it has nothing to do with tonnage, nothing to do with mech class, nothing to do with modules installed, nothing to do with premades, it merely puts each player in a match that gives them a 50/50 to win against others with similar win/loss records. And since according to PGI, all new players get a perfect median elo, they will face others with about a 50/50 win/loss percentage.

Now I can say in my last 100 drops in the last few days. I have had exactly 3 close matches. Or 3% that were within 2 mechs at the end. I have had 10 matches that were within 5 mechs but the majority at over 80% have had one side winning by more than 2 lances. So 8-12 mechs still alive at the end of the match for the winners and 0 alive for the losers. That is not anecdotal. That is fact.

And the matchmaker is to blame for lopsided matches. It is to blame because PGI does not take anything into consideration other than win/loss percentage. Which means the matchmaker currently being used by PGI is and never was ever designed to give even matches. In fact, an even match is an aberration by extension. Since the sole purpose of the current matchmaker is to attempt to make everyone fit into a bell curve of win/loss. And since Paul has posted on this subject twice in the last 30 days and announced, everything is working as intended. Then guess what. PGI is to blame for the broken matchmaker.

I hope that on Tuesday this changes.

What a matchmaker needs to have:

12 mechs on each side with a total tonnage value not more than 95 tons. Meaning neither side gets more than an Atlas (100 tons) advantage. Not much to ask for. But currently not part of elo.

12 mechs on each side with pilots near the same experience level. Meaning near the same amount of total drops. And I mean on a one for one basis. If one side has a pilot with 1k drops, then the other side should have one near that.

Premades exactly balanced on each side. If one side has a 4man, then the other side has a 4man. same for 3mans, 2mans etc.

ECM should never be more than +1 per side.

Arty, Air strike modules. Should never be more than +1 for one side.

Players with less than 50 drops should never be in a game with someone with more than 100 drops. Unless the new player dropped as part of a group.

Upgraded equipment meaning modules, engines, double heatsinks, etc should be within 20million cbills or about the value of a tricked out atlas for each side.

Now if you have all that in a matchmaker, you then get mostly balanced games. Premades face premades (but there should be a toggle for solo pugs that want to drop into premade games), No side out tons the other by more than a single atlas. No side has more than one more ecm than the other, no side has more than a small edge in arty/air, no side has all upgraded mechs vs. basic mechs. No side has mostly new pilots vs mostly veterans.

If a matchmaker can do that, then what are we left with? We are left with a game that leaves the players no excuses for win/loss other than pilot skill and teamwork. Which is exactly how a military simulation game should be. It should not have a matchmaker that attempts to manipulate win/loss percentage to make people have better self esteem. Or if you must have that, put it in the trial/training area for new players.

I don't play a simulation to have someone attempt social programs to help me with my feelings. I play simulations to push myself against other players. Not to drop into a game that has been intentionally loaded for one side to win and another to lose. More than 50% of the games I have dropped in feel this way. Like matchmaker attempted to give one side an easy win.

I just hope PGI fixes this. It has really caused a lot of players, new and old alike to leave the game or to just check in once in awhile. I was one of those that just got burned out by the horrific matchmaker. Calling it a matchmaker is a joke. It does not attempt to make an even match. Merely to turn every game into a bell curve win/loss 50%. Because it does not use any real data to attempt to even teams. NONE!

Otherwise you would not have a single screen shot where a team out tons another by 400+ tons. You would not have a screen where one side has 3 4mans and the other has none. You would not have a screen of one side with meta build hero mechs and the other with all basic mechs. Or a screen shot where one side has 8+ ecm mechs and the other side has zero. A real matchmaker would never allow that to happen, period! But I have seen all those screen shots, I have been on the winning and losing side of everyone of the above scenarios. Because ELO does not make even matches and is and never was designed to.

So make all the claims of statistical averages, not enough data. I can tell you a good matchmaker would not allow a single mismatch. And the current one does it time and time again. Meaning its broken (or working perfectly as intended according to Paul).

So I always ask myself, what was intended. PGI states what is intended for elo. To create a bell curve of win/loss and each game to assign a score to your player that puts you in matches where you have about a 50/50 chance to win. But what it does not do, is make an even match. It never was intended to. It never has except by accident.

My personal experience having played this game from about the second group of invites in closed beta. Is we had better matchmaking when it perfectly matched mech tonnage and we havn't had good matchmaking since. The best matchmaking I have seen was in the Marik Civil War. A community arranged matchmaker that even enforced no lrms/ecm on some maps and tonnage limits for each drop.

I just hope it changes soon. Because now, it sucks. Plain and simple. If I drop with a decent 4man we win 80% of our matches and most by stomps. If I drop solo, depending on tonnage of mech I choose and style of game. I can maintain a better than 50% win in hvy/assaults and less in lights. But that changes by game style. Lights and mediums are near wasted in skirmish unless dropping with premades. And once again, periodically elo will place 4-6 new players on one side against 2-3 premades. This guarantees a loss for oneside, every time. Saw this 5 times today alone. Was on both sides of it too.

So lets actually debate what elo is and isn't. It isn't a matchmaker. It is an attempt to manipulate win/loss so people do not feel bad. But it usually ends with one side stomping the other. I am literally friends with 100's of players that have been in since the beginning and are founders. It is unusual for us to see a close game and we comment as such on team speak or mumble or vent, whichever we are using. And for long term players like us, the unusual is the close game. The norm is the unbalanced game that ends in roflstomps.

Chris