152 replies to this topic

[Dudeman3k]

Sandpit, on 08 February 2014 - 05:29 PM, said:

FTFY

bahahaha, not exactly what I said, but sure. You can deflect it as you like.

[Rebas Kradd]

Dudeman3k, on 08 February 2014 - 05:26 PM, said:

Like I said before, PGI needs all the defending it can get. The only thing that really keeps me in is the fact its a MW title.... but I'm also keen enough to understand a bad game is a bad game.

It's an unfinished game, not a bad game. That alone invalidates a massive swath of the hate.

[Sandpit]

Dudeman3k, on 08 February 2014 - 04:59 PM, said:

Dude, Im just going to throw this out there. I've held my tongue long enough.

You are like Russ's lap dog, almost near blind to the fact that this game does in fact suck badly when compared to other titles. The only real appeal this game has over anyother is the fact it's a mech game, a BT title, and its F2P, but holy damn, this game sucks in terms of being a "game".

I read forum posts, sometimes I respond, sometimes I dont, but in every single one of them YOU are there... the junk yard dog of PGI. The one who is so blatantly blind to their incompetence in many things.

by all means, deffend them. To each their own, PGI needs more people like you....... but dude, you should really get out there and play more games to understand what being a gamer is.

Just to point something out. There was actually a discussion going on. On topic, constructive, no flaming, etc.

Until

Spoiler

you posted

[Mycrus]

Fact is pgi loosened mm elo difference so the I haz l33t crowd can stomp noobs.... all because jager something called russ out in a ngng podcast... so they adjusted it a bit but im still seeing wafflestomps... im pugging more now at least it is not a foregone conclusion that I will win in a 4-man...

my wlr has always been in the 1.4 range and then shot up to 1.6 with thousands upon thousands of drops that is a bit delta...

[Sephlock]

wwiiogre, on 07 February 2014 - 02:56 PM, said:

As it says,

50+ drops and now almost 95% roflstomps.

Stop dropping in light mechs ?

[Willard Phule]

Sandpit, on 08 February 2014 - 01:48 PM, said:

I'll go ahead and paste this in this thread as well as the other "MM sucks" threads I'm posting in

that 96%? They aren't experiencing stomps as often as the 2-4% in the extremes. So when you come up with an idea other than "MM sucks, PGI sucks, I'm uninstalling" that helps mitigate those stomps for the 4% that doesn't throw everythign out of whack for the 96% that AREN'T experiencing that, we might have something to talk about.

Meanwhile "PGI sucks, fix this!" and "Arggghhh I'm uninstalling because every game I play (which is a flat out lie) is a stomp!" does nothing to help figure out how to improve anything.


P.S.
There's no arguing against the bell curve figure I posted. It's a factual statistical truth. There's always that 2-4% at extremes (which is incidentally where we get the whole "There's always an exception to the rule") so please don't bother with the "Well that doesn't apply here because (insert reason why here)" MWO doesn't somehow magically bypass how statistics work.

So again, if you can come up with ideas on how to mitigate that 2-4% without screwing it up for the other 96% then lets hear it.

Sandpit, I agree that I can't argue with the statistics. I just question the statistics and where you got the data from. I know for a FACT that I'm not in the top 2%....as well as I know for a FACT that I'm not in the bottom 2% (let's face it, I know how to move forward and get out of 3PV...that puts me well above the bottom 2%).

So...if I'm in that middle 96%, why is it I end up on the losing side of a ROFLSTOMP 8 times to every win? Alternatively, if I'm in that middle 96%, why is it I consistently end up with people that ask questions like "how do I move forward?" or "what do you mean by lock targets? How do I do that?"

All good. We'll just agree to disagree on whether the matchmaker is broken or not. Personally, I think the whole concept of ELO is bad. It's overcomplicated and has been tweaked so many times, it's like shooting a deer rifle after your 4 year old fiddled with the scope.

As for "if you can come up with ideas on how to mitigate that 2-4% without screwing it up for the other 96% then lets hear it," how's this:

1) MANDATORY TUTORIALS. Not just this lame excuse for a tutorial that we've got right now. I'm talking about a Movement Tutorial, a Targeting/Weapons Tutorial and a Heat Management/Weapons Grouping Tutorial that are all REQUIRED to be passed (not just spacebarred through) before you're allowed to join a live match. This way, you'd be guaranteed that even if someone on your team is 'Brand New,' they at least know how to move and lock targets.

2) DO AWAY WITH THIS OVERCOMPLICATED ELO SYSTEM. Why does it have to be so [CENSORED] hard? Why not simply group by GXP Earned (go in 100,000 increments per bracket), then put together each team based on a BV (Battle Value) system? Going by GXP Earned guarantees that you're going to have 24 people with similar skill/experience levels together (that goes for new players as well as the elites). Introducing a BV system keeps the teams somewhat even. Sure, an Atlas D-DC is worth a lot of BV....but, if the guy slapped 3 LRM15s and no ECM on it, it'd be about what you'd see for a better put together mech. Granted, that sort of variable is going to differ depending on the experience of the person putting the mech together....but...that's sort of mitigated by the GXP thing, isn't it?

All good man. Nice roflstomp. GGClose.

[z00med]

THIS

MischiefSC, on 08 February 2014 - 03:59 PM, said:

Chris;

I get that you want to win more often. I really do. I get that you feel like there is some force, some evil group or person or entity or system or whatever that's keeping you down and preventing you from winning as much as you deserve to win.

I hate to say this but the problem isn't the matchmaker, it's you. This isn't intended as an attack or to be directly offensive but the issue you're putting forward is one of your personal perception and the psychology behind it.

The matchmaker does not seek you out to punish you. It works exactly the same for you, exactly the same, as it does for every single other player in the game. The matchmaker seeks to put you in situations where you're as likely to win as you are to lose, the only way to escape that is to get continually better. It's intended both to create a constant challenge and to prevent experienced players from constantly rolling new players. It doesn't care about you personally - nor should it. If you want to play in an environment where you win all the time go play a single player game.

Stomps are even more common in 12mans, in fact 12mans are almost universally a stomp. It's very rare to see a better than 12 to 2 outcome even if the relative skills of both teams were almost identical. It's not about 12 sets of 1 v 1, it's about organized teams using focused firepower.

In any case this isn't the first time you've had this discussion and the flaws and incorrect statements in your argument have been pointed out and addressed quite a few times. I could link them for you but there are 3 or 4 threads of them on the first page alone.

The short version - you're wrong, Chris, in saying the matchmaker is broken. Stomps happen regardless of matchmaking. Sometimes you play well, sometimes you don't. Sometimes your team does well, sometimes they don't. The matchmaker does what it's supposed to - try to keep things balanced in terms of weight and relative skill levels between players. Sometimes, due to availability of players within the needed weight classes, it has to go further afield in skill levels to fit tonnage and vice versa as it looks for an average but you're going to end up with around 50/50.

That you tend to take anecdotal extremes and blow them out of proportion is the issue here. Not the matchmaker.

has been the best read here EVER. 'nuff said.

I am kindly asking everybody who wants to whine about the evil matchmaker TO INCLUDE WELL DOCUMENTED (video from the last 30seconds, or at least both end screens as screenshots) DATA THAT CAN CLAIM STATISTICAL RELEVANCE (500-1000 matches should suffice). Otherwise the thread should simply be ignored. 10stomps in a row can easily be bad luck, so you should work on your zen instead of behaving like 3year olds who start to cry as loud as possible for not gettin sweets at the supermarket exit

Edited by z00med, 09 February 2014 - 09:21 AM.

[Too Much Love]

Previously I thought that considering small player base etc.creating good MM is a hard task, so PGI can be excused.

But after I saw the new Mechlab in UI 2.0 I realized that it's just a bad work. An ignorance.

The quality of the average match in MWO is awful. Sometimes it works clearly in the opposite way it should be working.

Ive been playing for a long time, so I know most of thd skilled players by names. I constantly see the situation when one team has all the good players, like two 4 men groups, and the other team consists only of ppl with the nicknames I see for the first time. MM could split this 2 4 men groups between the teams, but instead it put them on one side.

In majority of the matches I can predict the outcome of the game in the beginning by nicknames only.

I can, but MM can not.

Edited by drunkblackstar, 09 February 2014 - 09:31 AM.

[Thunder Lips Express]

PanchoTortilla, on 08 February 2014 - 01:36 PM, said:

Yet you all continue to give them your financial support.

Do you really expect them to change their ways when you reward them for taking a giant dump on your head every other week?

i love this game and i will continue giving them financial support until the time that i don't

[Nick Makiaveli]

Dudeman3k, on 08 February 2014 - 04:59 PM, said:

Dude, Im just going to throw this out there. I've held my tongue long enough.

You are like Russ's lap dog, almost near blind to the fact that this game does in fact suck badly when compared to other titles. The only real appeal this game has over anyother is the fact it's a mech game, a BT title, and its F2P, but holy damn, this game sucks in terms of being a "game".

I read forum posts, sometimes I respond, sometimes I dont, but in every single one of them YOU are there... the junk yard dog of PGI. The one who is so blatantly blind to their incompetence in many things.

by all means, deffend them. To each their own, PGI needs more people like you....... but dude, you should really get out there and play more games to understand what being a gamer is.

So not agreeing with you makes him stupid? Yep, now we know what side of the fence you are on.

Dudeman3k, on 08 February 2014 - 05:51 PM, said:

bahahaha, not exactly what I said, but sure. You can deflect it as you like.

I notice you didn't address the issues he raised in an earlier post, so looks like you are projecting.


Also, I can't help but notice people saying they lose 7/8 of their games seem to not realize they are the common denominator. If you aren't effective, then your team is down a man almost from the start. If you run off in a random direction and have a couple of newbies follow you, then you are down a lance.

Also, I like how Mycrus says he can win all the time in a pre-made, but occasionally loses in a PUG match....that's just e-peen stroking. Put it away...no one wants to see it.....

[PhoenixFire55]

Sandpit, on 08 February 2014 - 01:48 PM, said:

So again, if you can come up with ideas on how to mitigate that 2-4% without screwing it up for the other 96% then lets hear it.

Following your logic it should be 96% people happy with current MM. Yet somehow judging by the number of posts about MM and general in-game whine about MM I think its 96% people who are currently screwed. MM doesn't work, period. Random MM that we had in closed beta that only matched mech-to-mech classes for teams was producing far better matches on average.

Also this whole secrecy surrounding our Elo scores is just a pile of BS. If it really worked there would be no point in hiding our scores from us and yet they do. People been saying tonnage matching ever since bloody Elo-phase 1, and guess what, on April 15th (2015 I guess) they are gonna deliver same tonnage matching as their brilliant achievement.

[Sandpit]

PhoenixFire55, on 09 February 2014 - 10:20 AM, said:

Following your logic it should be 96% people happy with current MM. Yet somehow judging by the number of posts about MM and general in-game whine about MM I think its 96% people who are currently screwed.

I also explained why gathering any kind of data from the forums (statistically that is) is not going to give you accurate results. It's been officially stated a few times that a VERY miniscule portion of the player base even visits the forums. Not to mention that all of the complaint threads are basically the same people with a few new ones on each side posting.
That's also where the whole "duplicate post" comes into play.
I'd also like anyone to show me where I said the MM didn't need adjusting?

Also as far as the forums go, for ever person posting it's broken there's another posting it's not. This shows why you simply cannot rely on forums to determine statistical data

[Nick Makiaveli]

Sandpit, on 09 February 2014 - 10:56 AM, said:

I also explained why gathering any kind of data from the forums (statistically that is) is not going to give you accurate results. It's been officially stated a few times that a VERY miniscule portion of the player base even visits the forums. Not to mention that all of the complaint threads are basically the same people with a few new ones on each side posting.
That's also where the whole "duplicate post" comes into play.
I'd also like anyone to show me where I said the MM didn't need adjusting?

Also as far as the forums go, for ever person posting it's broken there's another posting it's not. This shows why you simply cannot rely on forums to determine statistical data

This. Plus do you really expect people like me who think that the MM is good enough for now while they tinker with tonnage limits etc to open a new thread just to say so?

Also, aren't you ignoring all the people who post in these threads saying it's good enough for now, but it's a good thing they are working on it since it needs work?

By your logic, if I posted a new thread every day about how SPL need to do 12 damage per shot, then eventually there will be 1 post per every 2 forum members, thus 50% of the population supports this change.... (point being look up how many threads wiiogre has started).

[Mystere]

PhoenixFire55, on 09 February 2014 - 10:20 AM, said:

Following your logic it should be 96% people happy with current MM. Yet somehow judging by the number of posts about MM and general in-game whine about MM I think its 96% people who are currently screwed. MM doesn't work, period.

I must be in your unscrewed 4% then. That's not bad for someone who has been playing solo all this time. As such, I must really be in Elo heaven.

PhoenixFire55, on 09 February 2014 - 10:20 AM, said:

Random MM that we had in closed beta that only matched mech-to-mech classes for teams was producing far better matches on average.

Have you already forgotten the whining that went on soon after(?) open beta in which people complained to the high heavens that they were regularly being stomped by 8-man teams. Or are you currently experiencing selective memory, possibly induced by trauma during that period?

Blame those very same whiners for the predicament people now face.

Edited by Mystere, 09 February 2014 - 11:14 AM.

[PhoenixFire55]

Mystere, on 09 February 2014 - 11:13 AM, said:

Have you already forgotten the whining that went on soon after(?) open beta in which people complained to the high heavens that they were regularly being stomped by 8-man teams. Or are you currently experiencing selective memory, possibly induced by trauma during that period?

Blame those very same whiners for the predicament people now face.

I do blame them. And as if now they don't whine? Same people same whine ... evil 4 man premades use em tactics and teamwork and kill me wuaaaaa. Its whine that brought Elo upon us, maybe same whine will get us rid of it. I never had any problems stomped by 8-mans back there, it was happening like once or twice out of 100 games.

[Joseph Mallan]

Nick Makiaveli, on 09 February 2014 - 10:05 AM, said:

Also, I like how Mycrus says he can win all the time in a pre-made, but occasionally loses in a PUG match....that's just e-peen stroking. Put it away...no one wants to see it.....

So what he is saying is when he is in a premade his team can cover his weaknesses better than PUG group. I'd say it makes perfect sense what he said. As in Obviously.

[wwiiogre]

Sandpit,

You are the one not actually showing any data to base your argument on. I have been collecting results for every game I play. Which means since ELO started I have 3500 drops of data. Including win/loss, k/d, etc, all of it. I keep track of every mech in each game, 4mans or apparent 4mans, 3mans, etc in each game. I keep track of my results when I drop in a light, medium, heavy, assault mech. I keep results for each type of game I play. I am a statistics junky.

And once again, I have an almost 2/1 win result over thousands and thousands of games.

My point is elo does what it is designed to do. It stacks games so one side wins and one side loses.

Until you start keeping track of your games and actually drop solo in lights in skirmish and then track the games. You have nothing to input into this discussion except blather and bull shart. The same is true for you as well Mischief. You have not done anything except come into this discussion and hurl insults. Which if you know anything about debating a point intellectually. The moment one side hurls insults, they have lost the debate.

I also know full well how statistics work. Which once again basing matchmaker on elo, when elo is purposely designed to force you into the median means it is not a true matchmaker nor is it how skill is based.

So until Sandpit and Mischief start tracking their games, show screen shots of how many games they played solo in lights, and show results for each style of game. You really have nothing to do in this discussion other than to hurl insults and obfuscate the point.

PGI's matchmaker does not match skill, does not match tonnage, does not match equipment. It matches nothing but a manipulated win/loss record. Which means no game played under it means anything. I have pointed out correctly that you can manipulate results and have a high win/loss by dropping in 4mans and by dropping in hvy/assault mechs. This is absolutely true, and is why you see so many lopsided wins. Mostly when one side out tons the other by more than 300 tons the game is a foregone conclusion. Now add in new players, 4mans and all the other things PGI does not use to balance a game and you get a majority of stomps and roflstomps. Conquest and sometimes assault allows the light tonnage side to snag a victory from the jaws of overwhelming tonnage advantage. But since PGI has manipulated conquest capture points on the big maps, that is less and less likely. Alpine is a good advantage with 4 conquest points less than 3 grid squares from each other.

And on, and on, and on.

ELO matchmaker is working exactly how PGI wants it to work. It does not make an even or balanced match. It does however force players with no knowledge of it to be forced towards the median of win/loss. Which means stomps and roflstomps.

I have showed how matchmaker could be made into an actual matchmaker. I have showed repeatedly how the current matchmaker is being manipulated by 4mans and heavy/assault cheese warrior online. Which is why people who know how to manipulate the matchmaker and take advantage of it want to keep on clubbing baby seals and will come here and argue, insult and refuse to debate the subject in an intellectual way.

If you are to slow to understand the above statement, what I am saying is idiots insult first in an argument because they cannot hold their own any other way. Con artists attempt to move the debate by distraction and by bringing points of discussion that have no bearing on the debate whatsoever.

Case in point, Sandpit posts a picture of a bell curve. What does that have to do with the fact that PGI in its own words has stated it is attempting to enforce a win/loss record near 50% for every player. That is in their own words. Multiple times. Which means matchmaker is not attempting to make balanced matches. It is attempting to make sure people with winning records have a better chance to lose.

What PGI has not done is make a matchmaker that actually balances sides in any way shape or form. Now that point has not been discussed by either Mischief or Sandpit. Please refute the above statement. You cannot, therefore you insult and distract. I have been personally insulted in this thread. The rules of conduct have been repeatedly broken in this thread by the naysayers.

Once again the Emperor has no clothes (or in other words PGI does not use a matchmaker that balances matches). That is my argument. Your responses have not touched on that basic premise.

Instead you insult and change the subject. My win/loss is not the debate. My wishing to win/lose a match is not the debate.

The debate is if even one match has one side with a 400 ton advantage matchmaker is broken.
The debate is if even one match has a side with a 4man and the other doesn't matchmaker is broken.
The debate is if even one match has a side with more new players than the other matchmaker is broken.
The debate is if even one match has a side with more expensive equipment by a margin of 100million cbills its broken.

I notice these facts are not disputed. Because you cannot dispute them they are facts. They happen in almost every single game. WHY? Because the most basic things you could possibly use in a matchmaker to balance sides are not being used by PGI.

They have said why they are doing it. That is also not in dispute. They have said they are trying to manipulate win/loss so that every player fits nicely into a bell curve. This is also not in dispute. When PGI has been confronted with overwhelming evidence that their matchmaker is completely broken they respond with and I paraphrase 'nothing to see here, move along, working as intended.' When Russ Bullock was on the recent ngng podcast and was asked specifically about matchmaker he said there would be improvement on the Feb 4th patch. He did not say there would be balance, he did not say it would be fixed. And he did not lie, they reset elo on the 4th so that, wait for it, everyone would be back inside the nice bell curve. That is what happened.

I would like a real matchmaker. I have not said I would quit playing. This is the best mechwarrior game ever made. PGI has got weapon, armor balance just about right. Eventually they even got ECM close to right. There are many things I think PGI has gotten wrong, but the core game is not one of those. Matchmaker is broken, consumable modules are broken and have nothing to do with this timeline or with lore but are nothing more than a money tap for PGI. They have made way to many small maps giving heavy/assault mechs the advantage. They have moved capture points on some of the large maps once again giving heavy/assault mechs the advantage. They have made the game less about role warfare and more about hvy/assault warfare. Skirmish mode added, for only one reason to make the hvy/assault players happy.

Last night I was in a game, randomly dropped and it was skirmish. I killed one mech, the rest of my team did less than 150 damage and all died. So we lost 12-1. Once again, multiple 4man teams on oneside, more new players in either trial or starting mechs on the other side. Tonnage advantage once again to the side with more 4mans. How much tonnage advantage you ask? They had 8 assaults, 3 heavies and a light. We had 2 assaults, 1 heavy (me in my K2). Was the match going to end any other way than a 12-1. Not really. No possible chance.

Now if this game was an aberration, and a statistical improbability. Then why did it happen over and over all night last night. Because once again, PGI's matchmaker takes none of the above into consideration. It only adds up your win/loss and attempts to adjust it towards median.

In my last game last night I played my Jaeger DD with 2xLBX10, 4xmg and 2xML. I killed 8 mechs had over a 1000 damage and we won. But I had to kill the other teams last 3 players. I wish I could say it was all skill, but in reality it was luck. All of them had been damaged and were cored. I was fast enough and used terrain well enough they couldn't snipe me so I rolled up on them, used advanced seismic and would wait at blind corners and see which way they were moving and then kill them at close range with lbx and mg fire into their open armor.

So yesterday I had one close game. It was the last and it left me with the feeling of why can't we have games that end close like that more often. Why can't that be the rule and not the exception. Not the fact that we won, but the fact that it was close at the end. Both sides were commenting during the last few minutes cheering and jeering. But it was close. It was fun and even had I died and lost it still was a great experience.

Of all the games I played that was the only close one. One other game on assault on alpine we won with 2 mechs left because the other team was so top heavy and slow they couldn't get back to their base to stop the cap. Yet they had 6 assaults left, a blackjack and a spider (which were chasing and killing me instead of stopping the cap, so my death and my buddies allowed us the win) which had they abandoned attempting to kill me and preacher would have won the game easily for the other side.

Before I started dropping in my Jaeger, I dropped in my new Ember for 5 matches and wrote down all the results. Tonnage, new players, trial mechs, end score, etc. Every game all 5 of them were stomps or roflstomps. There was not a close game with teams losing by less than 6 mechs. Even the games that were conquest or assault that ended in caps were never closer than 6 mechs on either side.

Then a friend joined me on TS3 and I put aside my data gathering, I kept keeping records but since I was not dropping solo, it kind of skews the results. We only had the one close game. The rest were stomps or roflstomps.

But since Sandpit and Mischief are not here to debate matchmaker and its merits and its problems. Which I have repeatedly pointed out. The fact it can allow such mismatches at all means it is broken. Unless of course you admit it isn't attempting to make balanced matches at all, but merely to manipulate everyone's win/loss record. Which is my point, yet somehow the detractors refuse to address this issue at all. Why?

chris

[Mystere]

PhoenixFire55, on 09 February 2014 - 11:20 AM, said:

I do blame them. And as if now they don't whine? Same people same whine ... evil 4 man premades use em tactics and teamwork and kill me wuaaaaa. Its whine that brought Elo upon us, maybe same whine will get us rid of it. I never had any problems stomped by 8-mans back there, it was happening like once or twice out of 100 games.

Then the whining is going the wrong direction. People should whine that 100% random drops be brought back. Also, use the KISS principle for justification.

[Joseph Mallan]

wwiiogre, on 09 February 2014 - 11:21 AM, said:

The debate is if even one match has one side with a 400 ton advantage matchmaker is broken.
The debate is if even one match has a side with a 4man and the other doesn't matchmaker is broken.
The debate is if even one match has a side with more new players than the other matchmaker is broken.
The debate is if even one match has a side with more expensive equipment by a margin of 100million cbills its broken.

chris

These are not correct though Chris.

Statistical perfection is 99.997%

That last 0.003% is your one game that there is a 400 ton mis match
and th eone match without a 4man...

Perfect is an illusion. You'd have a extremely good MM engine if you got a 400 ton mismatch 1 in 100. It would be pretty damb good at 95% of the time(that is an A grade performance).

[PhoenixFire55]

Mystere, on 09 February 2014 - 11:24 AM, said:

Then the whining is going the wrong direction. People should whine that 100% random drops be brought back. Also, use the KISS principle for justification.

Whine 'direction' doesn't matter. PGI will do their thing yet again, completely oblivious to whats going on. Just need to keep up the background hum that smth isn't right. Can't get any worse than now so its cool I guess.