540 replies to this topic

[KharnZor]

Sandpit, on 31 January 2014 - 11:33 AM, said:

The first 3 games
825/795 tons (825 wins 10-2)
935/830 tons (935 wins 10-2)
785/630 tons (785 wins 11-1)
720/875 tons (720 wins 12-6) numbers in () represent number of kills for each team

The tonnages were all decently close (except the last one and the third is kinda iffy in my opinion)

I noticed a few things. In the first three games charlie lance on the losing team performed abysmally. In one game they had a combined damage of less than 100
Charlie Lance Losing
1.)205
2.)759
3.)87
4.)838

Charlie Lance Winning
1.)667
2.)619
3.)745
4.)1111

Player's damage
1. 242
2. 33
3. 2
4. 533

There's a few issues here. First and foremost we have no idea the ELO or skill levels of the players involved. Secondly, we have no context for these numbers. Tailpipe, you yourself go from 2 damage to 533 damage.
In all of those matches the stomps only 1 champion (possibly new players?) in game 2 and only three in game 3.
There were several on both sides in game 4 but you were out tonned by nearly 300 tons (273 tons) and still managed to win.

Those 4 matches show nothing other than presumably the better team won and doesn't show us what caused those stomps. This is what I was getting at. Look at the stats for charlie lance in each of those. You have an entire lance that didn't do enough damage to take out a light mech. That's not a "premade vs. pug" or "elo out of whack within MM", that's simply bad play.

Even with all of the data breakdown that I just did it's still anecdotal. We don't know if you threw a couple of matches to help prove a point (or had others help you do this, not accusing just pointing out why anecdotal evidence proves nothing).

In the game where there wasn't a stomp there was nearly 300 tons difference in the two teams. That's a lot. So we know weight wasn't an issue. The claims of "i'm playing against high elo players out of my bracket" just simply doesn't work. It doesn't work because we have NO idea what our ELOs are, we have no idea what the grouping did to average the ELOs, we simply do not have the data to make statements like that.

I'm pointing all of this out not to say anyone is "wrong" but simply to say this is why anecdotal evidence doesn't work and to show that it may not be premades, elo, MM, etc. that causes a stomp. If that's the case then what is it? Sometimes it's one of these, sometimes it's simply player skill, sometimes it's just the simple fact that 2-3 derps on your team walked out and got caught in a bad spot (IE an entire lance doing less than 100 damage)

Good stuff.
You realize of course it wont make any difference. The derp thinks he's on a roll because he's posted some random screenshots so he cant possibly be wrong.

[Sandpit]

I'm just trying to put those screenshots in context of a bigger picture. Every day we see people posting "I got stomped" screenshots. Yet we don't have any data that explains why it happened. We have a few isolated games that show nothing more than the end result of a match.
We don't know the ELOs involved
We don't know the skill level
We don't know if a couple of players on a team just had a really bad match (happens to me all the time)
We don't know if 3 new players ran out and got killed in the first 30 seconds of the match

What we do know based on those screenshots is what I broke down and that's nothing more than weights, damages, and mechs used (we don't even know the loadouts so we can determine what weapons were used to put damage into context)
This is what those that say "Your example isn't a good one because it's anecdotal" are referring to. Stomps are bad but they're never going to be eradicated because sometimes they're simply caused by a couple of players playing like {Scrap} and that happens every day.
I understand the sentiment behind it, I'm just trying to point out that before you jump on the elo, mm, metagame, pug, or premade bandwagon, that you should know what exactly is causing the stomps.
That's why I disagree so vehemently with the solo queue ideas along with a few others about MM and elo adjustments. You can say all day long "well lets' try it and see" but the fact is that PGI has all of the data that DOES put games like those listed above into context along with every piece of data outside of "it was a stomp and these were the mechs used"
You can do one of two things really. you can either say
"I know I'm right even though I work on very small and poorly accrued data so they should make this change to fix it"
or
you can realize that PGI doesn't like the stomps anymore than we do and provide data without the commentary in a constructive manner so they can track it. Stomps = bad for everyone. They're not fun for either side (although there's always going to be that 10% that enjoy stuff like that because they're idea of a "good game" is when they roll over the enemies), but for the most part I honestly believe the population finds them boring and not what they're wanting in a game.

I just want people to understand that data doesn't always equate to accurate data.

[Ngamok]

@Sandpit

You do know some things. Charlie Lance is more than likely low ELO or new players with their default or close to default ELO. Not always true as my 3 man yesterday we were in Charlie because Alpha and Bravo were a 4 man each. You constantly see people in trial mechs in charlie.

[wanderer]

Let's put it this way. Perception is all.

All the player can see is losing 12-0, 12-1, 12-2...and nothing else.

They have no idea if one side had a premade and the other didn't.
They have no idea how badly damaged the other side was other than "dead" or "alive".

In short, the after-battle is really, really lacking on data that makes it appear like anything other than "you got your teeth kicked in, good stomp, you sucked".

The data is available. It needs to be shown, or perception is effective reality, no matter what "but really" bits you interject.

The player sees 4:1, 5:1, 6:1 and worse kill ratios and no other data? Why WOULD they think it was a fair fight, based on available facts?

[Sandpit]

wanderer, on 31 January 2014 - 02:14 PM, said:

Let's put it this way. Perception is all.

All the player can see is losing 12-0, 12-1, 12-2...and nothing else.

They have no idea if one side had a premade and the other didn't.
They have no idea how badly damaged the other side was other than "dead" or "alive".

In short, the after-battle is really, really lacking on data that makes it appear like anything other than "you got your teeth kicked in, good stomp, you sucked".

The data is available. It needs to be shown, or perception is effective reality, no matter what "but really" bits you interject.

The player sees 4:1, 5:1, 6:1 and worse kill ratios and no other data? Why WOULD they think it was a fair fight, based on available facts?

That's not the point at all. The point is finding the underlying problem. you can't fix something if you don't know what's causing it. Stomps will never be completely removed. There's simply no way to do that. You can, however, mitigate them. I'm not defending stomps, nor am I saying something shouldnt' be adjusted to lower the rate at which they occur. You also cannot eliminate stomps completely with a random MM at work. It's impossible. You could literally go through and get 900 damage and 5 kills and still lose because 3-4 players on your team discoed (intentionally or unintentionally), or they ran out full on rambo/leroy jenkins style, or high skilled players simply had a bad round. If a stomp occurs in a 12man that's ENTIRELY based on the players themselves.

Another thing of note is that no, PGI doesn't need to "share" their data. There's no reason for a player to see that data other than to curb their own ideas about "conspiracies" "PGI's stance on a situation", etc. Sharing the data would accomplish nothing other than paying someone on the PGI staff to collect and disperse the data to players for no other reason than to satiate their personal curiosities.

This is where you either have faith in PGI to work on the game and make corrections as needed or you don't. Nothing anyone is going to say is going to change a personal opinion on that. It's a moot point but it WOULD be nice to stop seeing players make assertions based on their personal experience and instead just post data without the opinion. That's how statistics and data collection works. Without bias, anecdotal, or personal opinion.

[Sandpit]

Ngamok, on 31 January 2014 - 02:14 PM, said:

@Sandpit

You do know some things. Charlie Lance is more than likely low ELO or new players with their default or close to default ELO. Not always true as my 3 man yesterday we were in Charlie because Alpha and Bravo were a 4 man each. You constantly see people in trial mechs in charlie.

You know Charlies lance is usually one of two things
1.) pugs
and/or
2.) the lowest elo on the team

That doesn't mean you know the elo, you just know that they're 1 or more points lower in elo than the above teams. The screenshots posted do NOT show a bunch of trial mechs. They showed 4 trial mechs over four games. It also does not indicate the amount of time a player has spent in the game. There are PLENTY of "vets" who have low ELOs. Again, that's what I'm trying to point out. Those screenshots do not put anything into context, nor do any opinions based on those screenshots.
Too many people see something like that and say "See? This proves I was right" when it simply doesn't work that way

Things I could incorrectly interpret from those screenshots
MM weight matching caused those stomps
Charlie lance being new players dropped in with high elo players caused it
Premades being involved caused it

Point being, that data does not in any way indicate what caused them to happen. The only thing it showed was that the player was involved in a few stomps. It doesn't even indicate how many matches were played total and how many were stomps. If we take the data that was provided at face value form the player, then 40+ matches = 4 stomps which is about 10% of the time.
Can we make that assumption? Not with any more accuracy than any other assumption being made based off of the anecdotal evidence provided by the screenshots and information given by the sampler

[PropagandaWar]

Lubalin, on 26 January 2014 - 10:12 AM, said:

When every round shows one team completely mopping the floor, something is wrong.

Answers about how ELO average really is equal in stomps, or how tonnage is going to be balanced at some vague point in the future aren't enough.

People are playing this game right now.

And it's broken.

We don't need new mechs, maps, or UI. Get the parts you've already built in working condition before you add new features. That's software dev 101. Stop designing new mechs and put your entire team to work on fixing the issues, or you'll keep losing customers.

There is something extremely fun about this game.

But I throw up my hands and quit every couple of months. I have no confidence at all in the people making this game. I know it's a labor of love. But maybe you guys need to bring in some outside help. Doesn't feel like you know how to fix this.

Why does this always happen to the best games? Sad.

Lub

Your crazy. Last night I was getting schooled. I had a bad night at first I thought it was just me, but then I noticed something I was in assault heavy matches with guys who more often than not just sat there. Hell you could tell them there was a lighter flank coming, but did they move? nope. Sorry but Some of the guys on a couple of them games I knew and if you just sat on your butt you were gonna get it kicked and it did. Back to the bad night. Last night was the first time ever in a year and a half that I got really frustrated. Did I go blaming others nope. It was my bad mood my bad day, my sucky game play. 10-12 games by the way are not that hard to turn around. More than that yeah its harder but not impossible.

[DaZur]

Sandpit, on 31 January 2014 - 02:40 PM, said:

There are PLENTY of "vets" who have low ELOs.

Why are you all staring at me?....

[wanderer]

Quote

This is where you either have faith in PGI to work on the game and make corrections as needed or you don't. Nothing anyone is going to say is going to change a personal opinion on that. It's a moot point but it WOULD be nice to stop seeing players make assertions based on their personal experience and instead just post data without the opinion. That's how statistics and data collection works. Without bias, anecdotal, or personal opinion.

I'd rather have more data and less opinion, and then maybe we'd have fewer biases, ancedotes in place of data, and opinions attempting to replace facts that are impossible to discover for lack of transparency.

Until such time, expect constant repeats of such threads up to the point at which players can SEE what actually happened in reasonable detail, because half-blind, ignorant (and ignorance is curable in players if you JUST LET THEM SEE WHAT HAPPENED), and frustrated players will take the limited data they DO get and infer the worst.

I don't want faith,I want results and I want initial attempts that at least show the forethought the Creator gave sand fleas. This matchmaker lacks in those regards and if the rest of 2.0 is like the cluster of fail that the UI is expressing, I -expect- nothing more than something worse than the previous.

[Sandpit]

wanderer, on 31 January 2014 - 03:13 PM, said:

I'd rather have more data and less opinion, and then maybe we'd have fewer biases, ancedotes in place of data, and opinions attempting to replace facts that are impossible to discover for lack of transparency.

Until such time, expect constant repeats of such threads up to the point at which players can SEE what actually happened in reasonable detail, because half-blind, ignorant (and ignorance is curable in players if you JUST LET THEM SEE WHAT HAPPENED), and frustrated players will take the limited data they DO get and infer the worst.

I don't want faith,I want results and I want initial attempts that at least show the forethought the Creator gave sand fleas. This matchmaker lacks in those regards and if the rest of 2.0 is like the cluster of fail that the UI is expressing, I -expect- nothing more than something worse than the previous.

You ARE seeing what's happening. You AREN'T seeing what's happening across the entire population and game. The problem? When someone posts something that says "Every game ends in a stomp isn't true"(or any other "this is true and factual for everyone in the game) you have a select few who want to jump up and down and say "you're wrong, here's a few screenshots proving it"
The problem is just because it happens to some doesn't mean it holds true for everyone in the game. That's why posting
"These are 4 screenshots from 47 games today that ended in stomps."
is MUCH better and useful than
"I'm posting these 4 screenshots because every game ends in a stomp and I have proof. It's because of elo (even though I don't know anyone's elo much less my own).

That's my entire point. You, as a player, don't need data just so you can rest easy at night knowing PGI is doing their jobs. That's like expecting McDonald's to release all of its statistical data about why they price McDoubles at $1 just so you know that's the best price point based on their data because you think it's too expensive because you live in an area that doesn't provide a job that supplies you enough disposable income to purchase them at that price. You'd be basing your "evidence" on a small section of the national data collected and base it on your small slice of the total.

They aren't going to do that simply because a customer "doesn't believe them when it comes to their decisions based on data they collect"

[Sandpit]

I'd rather have more data as well. That doesn't mean it is, should, or will ever happen. A lot of players seem to think they have some "right" to information in this regard. That's not how business works. You don't affect change by calling up the McDonald's customer service help line and screaming "YOU SUCK! YOUR PRODUCT SUCKS AND IT NEEDS TO BE CHANGED"

[Holding in your farts]

Sandpit, on 31 January 2014 - 03:55 PM, said:

I'd rather have more data as well. That doesn't mean it is, should, or will ever happen. A lot of players seem to think they have some "right" to information in this regard. That's not how business works. You don't affect change by calling up the McDonald's customer service help line and screaming "YOU SUCK! YOUR PRODUCT SUCKS AND IT NEEDS TO BE CHANGED"

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.

Edited by A banana in the tailpipe, 31 January 2014 - 03:59 PM.

[Sandpit]

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.

You missed the entire point.
You completely ignored every other post and statement made about your data. Those points stand and show why 4 screenshots doesn't prove "the matchmaker is borked and player's claims of "constant" stomps are true"
It also doesn't prove that the MM is the underlying issue. Once again, it doesn't prove anything other than over 48 hours and 40+ matches, you were involved in 4 stomps.....
GL&GH

[Marmon Rzohr]

A banana in the tailpipe, on 31 January 2014 - 01:48 AM, said:

Are you kidding me? What an apologist LOL!!!!!! "Just" 10 well placed shots... when 40+ alpha strikes obliterate trial mechs and Artillery/Air strikes one-shot people in the head are the norm. TEN "well placed" strikes might as well seem as far off as the land of OZ where your opinion is concerned.

I wasn't talking about 40 point alphas. I was talking about shots an average mech can make.
This means that if only two, three or even just one pilot played better between then the match would have gone differently. That's hardly the land of Oz...

Think about it.

One player tries really hard and lands couple of extra shots on a dangerous enemy mech like a highlander poptart. Now this highlander who would end the match with red armor and 800+ damage would be dead. Not only would one enemy be down, but one or two allies that mech would have torn apart would now live longer and themselves do more damage, resulting in perhaps one extra enemy mech down etc.

This is why matches snowball so easy. It is also why a single pilot can make a big difference with a few good shots.

This is the point I was trying to make. Many matches are close enough that they could have snowballed the other way based on the performace or positioning of just a couple of players. Even though the finally score would leave the impression of a steamroll.

(Some matches however really are not close, but that's much less common in my experience)

[Sandpit]

Marmon Rzohr, on 31 January 2014 - 05:12 PM, said:

I wasn't talking about 40 point alphas. I was talking about shots an average mech can make.
This means that if only two, three or even just one pilot played better between then the match would have gone differently. That's hardly the land of Oz...

Think about it.

One player tries really hard and lands couple of extra shots on a dangerous enemy mech like a highlander poptart. Now this highlander who would end the match with red armor and 800+ damage would be dead. Not only would one enemy be down, but one or two allies that mech would have torn apart would now live longer and themselves do more damage, resulting in perhaps one extra enemy mech down etc.

This is why matches snowball so easy. It is also why a single pilot can make a big difference with a few good shots.

This is the point I was trying to make. Many matches are close enough that they could have snowballed the other way based on the performace or positioning of just a couple of players. Even though the finally score would leave the impression of a steamroll.

(Some matches however really are not close, but that's much less common in my experience)

I notice this as well. Even if the end of match score is something akin to 12-3 that doesn't do the match justice if those 12 mechs on the winning aide are gimped and limping around waiting for a stiff breeze to finish them off, which is another reason stuff like end of match screenshots simply aren't good factual data.

[Nick Makiaveli]

A banana in the tailpipe, on 30 January 2014 - 11:37 PM, said:

So I tab back into the game and the very next match is another stomp. Still in my trial mechs at the lowest ELO possible. Sadly PGI has no control over how fast their game is spiraling out of control with casual players being driven off. I decided to go out of bounds rather than let the stompers take me.

You don't start at 0. Off the top of my head I can't call the number but I believe it's in the low middle range to start. Which pretty much kills your theory.

Ghogiel, on 30 January 2014 - 11:40 PM, said:

It's not irrelevent. No I'm not. And how do you even know that since you haven't a clue what margin for error the forum sample might have? It could be a 90% accuracy for all you know.

Ok what is the error of margin? Oh, you don't know either? You're right, it might be damn accurate. But we don't know do we? What we do know is that people who are happy just play the damn game. It's people who are pissed who are motivated to find a way to tell the company their thoughts.

Also, if we don't know how accurate it is, why should we listen to it? The devs have access to actual numbers, ie retention rates of new players, long time players, Founders etc. So pretty sure they have half a clue. Definitely a quarter clue, but probably a half.

Craig Steele, on 31 January 2014 - 12:16 AM, said:

Yup, your right.

It absolutely would require PGI to place a value on communicating better with the community and engage more resources, hire someone maybe. It's not about taking anyone away from any thing else, its about PGI putting more resources where there is value for them.

But as long as they perceive they can push forward what they do and it will be accepted by the market, then the incentive to do so is somewhat lacking.

They might be right, they might be making gazillions right now with the game / communication as is.

Pretty much. They are apparently ok with the way things are. I agree it could be better, but it is what it is.

Craig Steele, on 31 January 2014 - 12:19 AM, said:

This I am not so clear on.

If 5,000 people fill out a complaints form at their local Macdonalds, they all have an agenda? No one is going to say that they are the only 5,000 customer they had.

I just don't get why you would dismiss the volume of concern.

Apples and pipe wrenches. Go to eat, you have the complaint card in front of you. MWO, you have to leave the game, head over to a forum, figure out where to post etc.

Again, if you are upset at the restaurant you just fill out a card. Here you have to leave the game, log in to the forum, find/make a thread etc. Many people do. Add in the happy people, how many are going to the trouble (for a non-forum user it is a hassle) to post?

Also, no you wouldn't say those are the only customers they had, they are going to check and see they had 350,000 customers.

So tell me, how many people play MWO? What percentage use the forums?

Exactly. We don't know. We just know of the people who post here. It isn't reliable data. Therefore any conclusions that come from it are suspect.

It's not personal, it's just facts.

[Wispsy]

A banana in the tailpipe, on 30 January 2014 - 11:37 PM, said:

So I tab back into the game and the very next match is another stomp. Still in my trial mechs at the lowest ELO possible. Sadly PGI has no control over how fast their game is spiraling out of control with casual players being driven off. I decided to go out of bounds rather than let the stompers take me.

You are not the only one complaining that people like me get placed against fresh accounts these days...Although Peepers does not play assaults much at all so he might have been our weight! This is how the vast majority of my games over the past couple of days look like, albeit with multiple different mechs in group, it was especially unfortunate of you to catch us at the end of the night when we went randomly heavy for our final drops...

On a side note, most of the people in this thread are right, 12-2 match scores can indicate very closely matched teams, it is just very easy to snowball. Once you have a 3/4 mech advantage unless a lot of cored people get picked off very quickly you are simply going to overwhelm them and anybody too hurt can just hang back and put in only the occasional shot from safe places.

Edited by Wispsy, 31 January 2014 - 07:19 PM.

[Wispsy]

Ngamok, on 31 January 2014 - 02:14 PM, said:

@Sandpit

You do know some things. Charlie Lance is more than likely low ELO or new players with their default or close to default ELO. Not always true as my 3 man yesterday we were in Charlie because Alpha and Bravo were a 4 man each. You constantly see people in trial mechs in charlie.

Aye I am pretty sure Elo has nothing to do with it, simply the lower down you are the smaller your group, with people placed randomly to fill gaps. I have seen Wolfens sitting in the top of Bravo lance before above some much higher skilled and (I would bet real money on this) Elo in Charlie.

[KharnZor]

@ Wispsy, did they finally impound your spider? never seen you in anything but that evil arachnid

[Holding in your farts]

Nick Makiaveli, on 31 January 2014 - 06:44 PM, said:

You don't start at 0. Off the top of my head I can't call the number but I believe it's in the low middle range to start. Which pretty much kills your theory.

I don't have a "theory" anymore... I went out and proved it. You and the others like Sandy are full of talk where I actually took action and did the research. You're welcome and I've got a cigar fired up with a glass to scotch to celebrate.

Wispsy, on 31 January 2014 - 07:14 PM, said:

You are not the only one complaining that people like me get placed against fresh accounts these days...Although Peepers does not play assaults much at all so he might have been our weight! This is how the vast majority of my games over the past couple of days look like, albeit with multiple different mechs in group, it was especially unfortunate of you to catch us at the end of the night when we went randomly heavy for our final drops...

You're the guy who says "death awaits you" or "death comes for you" ect correct? Yeah I was placed in a stomp vs you while my cadet bonus was still intact. Just more proof the matchmaker and this game is screwed without PGI doing something major to increase the player base.

This is one instance where it sucks to be right, and nobody can refute it despite how hard they claim ELO, the matchmaker, or even PGI themselves work. The population cannot support this style of matching and PGI has 3 options:

1. Some sort of incentive like free MC so players will come back / give MWO a try.

2. Abandon MMO style servers and let players host their own servers so they can segregate high and low skilled players, or at the very least allow players to move on from a server they are outclassed on instead of getting placed in stomp after stomp wasting the player's time.

3. Allow the game to continue spiraling out of control and circle the drain.

Game, set, match. Checkmate ladies... stick a fork in this one.

Edited by A banana in the tailpipe, 31 January 2014 - 10:06 PM.