49 replies to this topic

[Mystere]

Cosmocat, on 25 April 2015 - 04:47 PM, said:

So the idea of the matchmaker being responsible for all of the stomps we experience is something that is put forward so often that it's basically become an accepted fact in the Mechwarrior Online community, but (surprisingly to me) it isn't suggested very often that lopsided results might just be the normal outcome of 12 vs 12 matches, even when the teams are evenly matched.

Goebbels would be proud.

[MrMadguy]

You're lucky, man. Cuz for me it usually something like 1 win vs 10 loses. This have been happening with me since the implementation of 3/3/3/3 system, which extremely narrowed range of players, available for matchmaking. And about a week ago MM's stupidness came to it's apogee - I've got only 4 wins during 12 hours of playing (with breaks), I earned only 2M CBs and my W/L dropped by 2% solely during this time, lol. So, your situation isn't the worst one. It could be much worse.

[Summon3r]

i have found in the last 2 months that it has gotten exceptionally worse, not sure if a patch messed with the ELO/MM system but sometimes its like dropping with guys that are in there first few drops or something

[RAM]

Kjudoon, on 25 April 2015 - 07:04 PM, said:

Don't forget. Elo is trying to keep you at a 50% victory percentage.

Not how ELO functions.

Cosmocat, on 25 April 2015 - 04:47 PM, said:

So the idea of the matchmaker being responsible for all of the stomps we experience is something that is put forward so often that it's basically become an accepted fact in the Mechwarrior Online community, but (surprisingly to me) it isn't suggested very often that lopsided results might just be the normal outcome of 12 vs 12 matches, even when the teams are evenly matched.

How ELO functions.


RAM
ELH

[MrMadguy]

Summon3r, on 25 April 2015 - 10:38 PM, said:

i have found in the last 2 months that it has gotten exceptionally worse, not sure if a patch messed with the ELO/MM system but sometimes its like dropping with guys that are in there first few drops or something

I guess, this is due to devs trying to push CW hardly during the most recent time - most events were CW-oriented for example.

This is really bad for me, cuz they have implemented new exclusive mode without finishing development of previous one and without being able to fill both modes with players. And I personally don't play CW due to some reasons, which means, that if they'll continue pushing it so hard and killing Group Queue as the result - I will have to quit this game. Major reasons, why I don't play CW:
1) CW has objectives and I think, that objectives are forcing players to do weird things instead of actually playing the game.
2) CW - is purely group content and I don't play in a groups, cuz due to some reasons I can't take virtual social obligations on me on the top of real life ones. CW requires coordination, cuz otherwise stupid maps, lacking cover and hidden ways to came into close range with other team are turning CW into shooting gallery.
3) CW - is the way to force playing Light/Medium 'Mechs, devs were searching for, in order to increase sales for this unpopular 'Mech classes. And I don't play Light and Medium 'Mechs. I personally think about 'Mech classes in MWO as character classes in MMOs - some classes suit your playstyle, some don't. Heavies - the only class, that suits my gamestyle. I also play Assaults, because "it's cool". Forcing my to play Lights and Medium 'Mechs in MWO - is the same terrible idea, as forcing players to have characters with all classes in MMOs.

Edited by MrMadguy, 25 April 2015 - 10:59 PM.

[Deathlike]

I don't have answers to this, but there's usually 3 things contributing to this (I can't stop using 3, help?)

1) State of mind

This sounds cheezy, but over time when you're losing (or particular when you're tired), many times, bad decisions start being executed... and what's worse is that it can get in your head.

One of the things I actually did for those solo tournies was to simply close down the game, do something else, and then come back.

As much as it is anecdotal evidence, the biggest issue is that it can be a mental thing... even if it isn't getting a 1 kill a match, or just making sure you're not getting killed early in a match.


2) Elo

The more you play in a group AND win, the more likely your Elo will be out of whack in solo... and it's usually significant.

The thing is though that as you succeed in groups, the group constitution is not usually the same in the ones you win in and the ones you lose in. It is possible that the group can have a streak of losses... in most cases the "combined" Elo in the group (group size increases your Elo weight) would alter how much carrying it has to do on the whole.

If you think of it like "inserting some small elite unit" on your team, that would have a similar effect to "inserting just a bunch of guys". While the group Elo (through whatever magical Elo weighting) isn't going to be reflected perfectly, things get amplified.. not just the players, but the quality of play you are getting.


3) Playerbase size - yes, it still matters!

Depending on the time of day you are playing, regardless of group or solo... you will see a lot of the same groups/players playing in the matches. While this is somewhat of a temporal issue (people in the queue finishing the match around the same time you do, has a great chance of getting them again, occasionally seeing them on the opposing team), it is actually a lot more indicative of the playerbase size as a while. If the playerbase activity is even 3 times what it is now (excluding the event, because that is actually exacerbating the issue), then this wouldn't be such a regular thing.

This becomes really obvious in a solo tourney/event, and the probability/outcomes don't lie. Of course, this assumes people try to drop ASAP, instead of occasionally waiting for something to put together a new build and the launch. So, it is what it is.

I wish I could suggest better things, but this is a reality in MWO. I wish I had better news, and while you can feel free to disagree with my points, I've play enough games solo or group to pay attention to nuances of the MM, regardless of the rules that have been instituted for it (3/3/3/3 with release values and all). These things tend to allow such streaks to happen, and certainly a major part is the player's skill. However, if player skill is consistent (and the person doesn't get massively influenced from match to match), then a lot of these things listed about tend to contribute to the kinds of effects listed AND occasionally get magnified by people that get really ornery about their losses (mind you, I feel it occasionally myself).

So, it is what it is... it's also PUG life.

[Sarlic]

Lily from animove, on 25 April 2015 - 05:16 PM, said:

randomness, and not so randomness randomness

pay attention if after such a match the majority of the players is the same at the next matchup.
Because especially in CW the way how drops are filled can provide you often with the same people sicne they wikll very likely exit the game and enter the together again..




That idea is dangerous. because does that mean i have to "unlock" hardmode? meaning I can stay midmode and wreck casuals? hardmode should probably apply at a k/d of hmm, lets say 1.8 but then even this can eb faked losing some matches on intention.

basic ideas, there are many. But hardly any idea is safe from being abused and exploited by people seeking easy games for trolling, stomping or epeen stroking.

You have to unlock each mode. You can't stay midmode, because when you play that mech you gain experience. After a goal is hit, it gets unlocked. You can't switch back to the old que when it gets unlocked.

For example you just bought a Spider. Spider needs three goals before you can join the veterans. That's almost (insert numbers here) of GXP. New players que only.

Your main chassis is a Thunderbolt. Everything unlocked, mastered and so far good outfit. That one already has unlocked all 3 goals and can only play with the veterans, it has no use to mixup with new players.

It's not based on KDR but based on GXP and perhaps another factor such as playtime or something.

I know alot of systems can be abused and im against it, but the madness we have now with new players mix match avarage and high suited players isnt much pleasant either. Atleast by splitting it would be atleast workable for the higher players and hoping to see more teamwork.

Well... Atleast it's a basic idea for the solo que.

Edited by Sarlic, 25 April 2015 - 11:53 PM.

[Taffer]

Yea, you can really feel the lose streak when you are on it. It LASTS. I can go into a match and say without a doubt that my team will lose because I'm on a lose streak. Sometimes I announce it just for fun, and it's true like over 90% of the time. Doesn't matter if I carry hard or anything, it's destined in some bizarre way. Like the game KNOWS I'm not going to win and that's why it put me there.

I know it sounds paranoid and silly, but I really think it's a thing. It's not just lose streaks as in, "oh hey I happened to lose 14 times in a row gee wiz." It's more like, "oh **** I lost 7 times in a row, I know for a fact I will lose several more.

talking about solo queue btw

Edited by Taffer, 26 April 2015 - 12:09 AM.

[MrMadguy]

Taffer, on 26 April 2015 - 12:08 AM, said:

Yea, you can really feel the lose streak when you are on it. It LASTS. I can go into a match and say without a doubt that my team will lose because I'm on a lose streak. Sometimes I announce it just for fun, and it's true like over 90% of the time. Doesn't matter if I carry hard or anything, it's destined in some bizarre way. Like the game KNOWS I'm not going to win and that's why it put me there.

I know it sounds paranoid and silly, but I really think it's a thing. It's not just lose streaks as in, "oh hey I happened to lose 14 times in a row gee wiz." It's more like, "oh **** I lost 7 times in a row, I know for a fact I will lose several more.

talking about solo queue btw

Yeah. During my 12 hours lose streak I was getting fun via telling my team, that they're going to lose, cuz I'm in their team, too. Being able to predict result of the match with 99% accuracy - is the only fun thing about having lose streaks. This is what I call "we call it RNG, but in fact it's just thing, you can't have control over" - next and even worse degree of "we call it RNG, but in fact it's just thing, you can't predict" gamedevs' unfair "RNG" abuse trick.

When you're on a lose streak, you actually can notice it via the way map and spawn points rotation is working for you. You constantly get the worst map for the 'Mech, you've chosen - hot maps for laser/PPC build (Caustic in most cases), large maps without cover (Alpine, lol) for brawler builds, small maps for sniper builds (River City). Also you may notice, that if you're using the same 'Mech - you have increased chance to get the same map several times in a row. Got Caustic for PPC build 'Mech, then other map for other build and then again Caustic for same PPC build 'Mech. And finally - spawn points. You may notice, that you're getting the same spawn points again, again and again. Always lower left spawns for Assault 'Mech at Alpine - worst spawns for Assaults ever, cuz it takes up to 5 minutes of walking to group up with your team.

As always, this is the second game in a row, where developers are abusing "RNG" to get profit from it. In the past I also experienced such a terrible things, like manipulating players' loot in MMOs in order to artificially stretch the content. Devs are abusing "RNG" cuz players can't prove anything in this case. And it works in most cases, if it's used carefully. But sometimes devs are becoming too greedy and start feeling unpunished. And that's rising degree of abusing of so called "RNG" to enormous heights. Something like event with 15% chance to happen per try not happening during 200 tries or the same loot (2 items from first two bosses and 2 items from last boss) dropping from 3 bosses in instance 3 times in a row.

So PGI isn't the first company, who abuses so called "RNG" in order to manipulate players' chances to win. And not the last, I guess.

Edited by MrMadguy, 26 April 2015 - 01:49 AM.

[Milocinia]

The problem, that I'm lead to believe, is that this "ELO" thing is primarily designed for 1v1 matchmaking.

Every player that might be in the same ELO "bracket" as you can have completely differing playstyles and abilities. Then you have mech preferences, favourite maps and so on.

There are so many variables which ends up making your ELO score thingy completely meaningless and the resulting matchmaking totally derp.

But according to Russ's metrics, it's all working as intended. Re: Annoying the complete **** out of a majority of the playerbase.

[El Bandito]

I am gonna add this thread on my "Why 8v8 is better for Solo-Q" list.

Edited by El Bandito, 26 April 2015 - 03:04 AM.

[anonymous161]

idea of doing 20 matches in cw to get a decent reward gives me a headache no way I'm investing that much time into the game at once.

[Bishop Steiner]

MrMadguy, on 26 April 2015 - 01:03 AM, said:

Yeah. During my 12 hours lose streak I was getting fun via telling my team, that they're going to lose, cuz I'm in their team, too. Being able to predict result of the match with 99% accuracy - is the only fun thing about having lose streaks. This is what I call "we call it RNG, but in fact it's just thing, you can't have control over" - next and even worse degree of "we call it RNG, but in fact it's just thing, you can't predict" gamedevs' unfair "RNG" abuse trick.

When you're on a lose streak, you actually can notice it via the way map and spawn points rotation is working for you. You constantly get the worst map for the 'Mech, you've chosen - hot maps for laser/PPC build (Caustic in most cases), large maps without cover (Alpine, lol) for brawler builds, small maps for sniper builds (River City). Also you may notice, that if you're using the same 'Mech - you have increased chance to get the same map several times in a row. Got Caustic for PPC build 'Mech, then other map for other build and then again Caustic for same PPC build 'Mech. And finally - spawn points. You may notice, that you're getting the same spawn points again, again and again. Always lower left spawns for Assault 'Mech at Alpine - worst spawns for Assaults ever, cuz it takes up to 5 minutes of walking to group up with your team.

As always, this is the second game in a row, where developers are abusing "RNG" to get profit from it. In the past I also experienced such a terrible things, like manipulating players' loot in MMOs in order to artificially stretch the content. Devs are abusing "RNG" cuz players can't prove anything in this case. And it works in most cases, if it's used carefully. But sometimes devs are becoming too greedy and start feeling unpunished. And that's rising degree of abusing of so called "RNG" to enormous heights. Something like event with 15% chance to happen per try not happening during 200 tries or the same loot (2 items from first two bosses and 2 items from last boss) dropping from 3 bosses in instance 3 times in a row.

So PGI isn't the first company, who abuses so called "RNG" in order to manipulate players' chances to win. And not the last, I guess.

While I think you bring up some interesting points, about the spawn, map selection and all that, those actually are likely facets used by MM to try to hit that golden ratio. I had noticed some trends, but now I really want to chart it.

I would say that Russ has never hidden the intent with Elo. It is, going to try it's best to keep you at a 50/50 win loss, and shockingly, when polled that is pretty close to what most people end up with, exceptions being the uber leet and certain teams that are composed of such.

Even bads tend to end up there by being carried. And in fairness, results seem better than pre Elo.

No, mu complaint isn't so much Elo trying to get you to 50%, or the way it has to try to balance team Elos, with some wildly disparate results. It's more the part of the algorithm that decides when to punish, when to reward. It's not losing about half the games that bugs me (well, it does, but so be it) but the way it does it, where instead of giving you small bites either way, a few matches here, a few there, good and bad, to balance your Elo, it's those long 20 game streaks.

Those simply suck the joy out of playing, and make me turn off the game and watch Netflix instead.

And folks can say carry harder all they want. Unless you are Heim or Koreanese, etc, there ain't a lot you can do but try to go out in a blaze of glory when you consistently get 11 derps assigned you, and you find it to be you vs the other team 12-12 vs 1, in the first 2 minutes.

Darth Bane001, on 26 April 2015 - 02:44 AM, said:

idea of doing 20 matches in cw to get a decent reward gives me a headache no way I'm investing that much time into the game at once.

rate of drops I got for CW...that'd be 24 hours of play.

[MrMadguy]

Bishop Steiner, on 26 April 2015 - 02:45 AM, said:

While I think you bring up some interesting points, about the spawn, map selection and all that, those actually are likely facets used by MM to try to hit that golden ratio. I had noticed some trends, but now I really want to chart it.

I would say that Russ has never hidden the intent with Elo. It is, going to try it's best to keep you at a 50/50 win loss, and shockingly, when polled that is pretty close to what most people end up with, exceptions being the uber leet and certain teams that are composed of such.

Even bads tend to end up there by being carried. And in fairness, results seem better than pre Elo.

No, mu complaint isn't so much Elo trying to get you to 50%, or the way it has to try to balance team Elos, with some wildly disparate results. It's more the part of the algorithm that decides when to punish, when to reward. It's not losing about half the games that bugs me (well, it does, but so be it) but the way it does it, where instead of giving you small bites either way, a few matches here, a few there, good and bad, to balance your Elo, it's those long 20 game streaks.

Those simply suck the joy out of playing, and make me turn off the game and watch Netflix instead.

And folks can say carry harder all they want. Unless you are Heim or Koreanese, etc, there ain't a lot you can do but try to go out in a blaze of glory when you consistently get 11 derps assigned you, and you find it to be you vs the other team 12-12 vs 1, in the first 2 minutes.


rate of drops I got for CW...that'd be 24 hours of play.

It's already mentioned before and I agree with it - ELO is only suitable for 1vs1 or constant team vs constant team scenarios. The main intention of ELO is not to achieve ~1 W/L ratio, but to provide most enjoyable gameplay via providing allies and enemies with proper level of skill. And in this case ELO should also take into account personal performance - not only average Team performance. In easiest case in should also try to achieve ~1 K/D ratio (match score or some anther personal performance measure method in more complex case) via controlling average ELO level of both teams itself - not only difference of ELO levels. Some players, like me, have <1 K/D and some players have >2 K/D. If your performance is too good - MM provides allies and enemies with higher average ELO levels. If your performance is low - with lower average ELO levels. And you won't carry bad team and won't be stomped via superior enemies. For example sometimes my team wins the match, but I feel like I haven't affected this result at all - for example was one-shotted on the first minute of the match. That means, my team average ELO level was much higher, then my own - I should have been put into team with lower average ELO against enemies with lower average ELO. Currently MM works exactly the opposite way, which causes constant roflstomps instead of close matches.

[Wintersdark]

Oh, god.

Deathlike already covered the steps to mitigate issues.

There's so much wrong in this thread about how the MM works, it's almost physically painful.

We know _exactly_ how it works. The math, how teams are constructed, the algorithm.

The matchmakermatchmaker via Elo doesn't try to get you to a 50% WLR. That's a side effect of its working, but not its goal. This is important to understand because believing the former leads to making stupid assumptions like "The matchmaker set this game up for me to lose".

It leads to Special Snowflake Syndrome, where players start believing that they're somehow important to the matchmaker. You are just one number out of 24 numbers, no more or less important than anyone else.

The matchmaker doesn't care about your load out or mech(beyond weight class).

Mr. Madguy above is a good example of Deathlike's first point, and basic " how to win" training. Believing your going to lose will heavily impact your play. Telling people that will impact theirs. Whether they think your just throwing the game, or they also believe the MM is a malignant monster trying to make them lose, that belief has an impact.

Quote

For example sometimes my team wins the match, but I feel like I haven't affected this result at all - for example was one-shotted on the first minute of the match. That means, my team average ELO level was much higher, then my own - I should have been put into team with lower average ELO against enemies with lower average ELO. Currently MM works exactly the opposite way, which causes constant roflstomps instead of close matches.

getting "one shotted" at the start of the match doesn't say anything about your Elo vs. Your team average. Nothing at all.

But you know what? You had a HUGE contribution to how that match turned out, not no contribution.

[Zookeeper Dan]

I don't know if it's the matchmaker or my state of mind I'm just glad to hear other players have long losing streaks too. The other night I had ine win followed by 8 loses. I should have stopped, but I really wanted to go out on a win so I kept playing "one more game".

Although it makes me feel better to blame the matchmaker.

In reality it's probably random chance. Flip a coin enough times and you will have streaks of heads or tails. We just remember the streaks of losses.

[Mawai]

Sigh. Not this again.

Honestly, it likely has absolutely nothing to do with Elo.

In a PUG match a certain number of players run in random directions or just do their own thing. These folks usually either die first by running into the opponents or last by not helping the group that does run into the opponents.

Whichever side loses these folks first stands a much better chance of losing the match because they will be down 2 to 4 mechs almost immediately.

I see folks on BOTH sides doing this. If you are unlucky then the ones on your team die and if you are lucky it is the ones on the other team.

Once your side is down a couple of mechs it can be very easy for the game to steam roll from there ... especially if your team is not together and the opposition encounters several mechs. They roll over them and your fire is not focused and usually ineffective.

It has nothing to do with Elo .. it has to do with team work. A few words in chat or a few brief comments on coms can often get a PUG group working a bit more cohesively ... do that and your odds of winning probably go up 25%. Don't do it and you are depending on someone on the other side NOT doing it and the luck of the derp to fall your way as folks wander about randomly.

Some folks are in your Elo range because they aim well ... some are in the same range because they stick together and work well on a team. Get too many folks with good aim and bad team work ... they wander all over and die fast ... if you have the choice (which you usually don't) take team work over aim.

[KodiakGW]

But Bishop, I've been told there is nothing wrong with Elo and MM. People have been typing up page long arguments against my post saying nothing is wrong.



Yeah, right. Top damage, losing team, again, not even basic leveled recently purchased JM6-FB.

Bishop Steiner, on 26 April 2015 - 02:45 AM, said:

Those simply suck the joy out of playing, and make me turn off the game and watch Netflix instead.

Yep, and for me, Defiance. Hey, Remo Williams is back on Netflix. Now if they only would add Big Trouble in Little China and Buckaroo Banzai, you could have a cheesy 80's movie marathon.

Edited by KodiakGW, 26 April 2015 - 07:38 AM.

[LordKnightFandragon]

KodiakGW, on 26 April 2015 - 07:36 AM, said:

But Bishop, I've been told there is nothing wrong with Elo and MM. People have been typing up page long arguments against my post saying nothing is wrong.



Yeah, right. Top damage, losing team, again, not even basic leveled recently purchased JM6-FB.



Yep, and for me, Defiance. Hey, Remo Williams is back on Netflix. Now if they only would add Big Trouble in Little China and Buckaroo Banzai, you could have a cheesy 80's movie marathon.

I dont understand how people can claim nothing is wrong when 98% of games end just like yours....4-12......and never better....that is a stomp......if it was more like 7-12, that was just a fairly bad game, but 4 or less to 12...that is an ass kicking.

I mean, look at your side match scores, you alone got your entire team's combined score...or just about.....yet for all your effort, you lose. Maybe its starting to click in peoples minds that it doesnt matter what 1 player does, all 12 need to give it a go for it to be a good game. Carrying doesnt even happen. Carrying is just 1 player who managed to be in the right place at the right time in a single game, they are 1 in 100 games and only possible if the rest of your side manages to participate and accomplish sufficient damage to allow you to exploit it......

Edited by LordKnightFandragon, 26 April 2015 - 09:23 AM.

[Deathlike]

Wintersdark, on 26 April 2015 - 04:49 AM, said:

Oh, god.

Deathlike already covered the steps to mitigate issues.

There's so much wrong in this thread about how the MM works, it's almost physically painful.

We know _exactly_ how it works. The math, how teams are constructed, the algorithm.

The matchmakermatchmaker via Elo doesn't try to get you to a 50% WLR. That's a side effect of its working, but not its goal. This is important to understand because believing the former leads to making stupid assumptions like "The matchmaker set this game up for me to lose".

It leads to Special Snowflake Syndrome, where players start believing that they're somehow important to the matchmaker. You are just one number out of 24 numbers, no more or less important than anyone else.

The matchmaker doesn't care about your load out or mech(beyond weight class).

Mr. Madguy above is a good example of Deathlike's first point, and basic " how to win" training. Believing your going to lose will heavily impact your play. Telling people that will impact theirs. Whether they think your just throwing the game, or they also believe the MM is a malignant monster trying to make them lose, that belief has an impact.

getting "one shotted" at the start of the match doesn't say anything about your Elo vs. Your team average. Nothing at all.

But you know what? You had a HUGE contribution to how that match turned out, not no contribution.

The belief in "I'm/We're going to lose" is generally detrimental to play. It just is. Usually the player tends to run into the "self-fulfilling prophecy", in which case it then requires sheer luck before a win happens.

One of the worst things to do in coms is to say that out loud. Although, I guess it's in the nature of some people, but it must be a reminder to all people that you actually must "do your part" in NOT being the reason for your team's loss. In the occasional chance that you DO win vs a significant force (Lords or whatever elite team of your choice), it's because YOU DIDN'T THINK about who you were playing... rather, you thought more about HOW YOU PLAYED.

This is first and foremost rule about success in MWO (or any game for that matter). If you cannot objectively look at your overall game, it's hard to figure out how you can do better. If you were the first to die in a match, it is actually possible that YOU did something wrong.

It is a hard way to determine whether you'll ever get better at this game... and as long as you don't get down on yourself and only worry about how you need to do everything to succeed, you'll be better for it.

False confidence in success can easily be problematic (for self-evaluation) as it is for prophetic projections in failure. If you stop worrying about premades and solos and focus more about working together, the game and YOU would be better for that.

Sadly... it is hard for some. I could probably do a commentary on all the games I've done (or watching others), but ultimately it's not hard to see success and failure if you don't even do the job well.

Edited by Deathlike, 26 April 2015 - 09:54 AM.