104 replies to this topic

[Clay Pigeon]

There's no such thing a cheese. There are effective builds, and ineffective builds, and some middle ground between them. You were in a SRM boat a1, you just got out played by our A1.

[Commander Kobold]

EgoSlayer, on 26 February 2013 - 08:49 PM, said:

And the previous match making prevented these exact same scenarios how?

If 'Cheesebuilds' are so overpowered these player's ELO will rank them up to the point where you won't see them anymore and you'll be matched against other players who are not running these builds. It becomes self correcting.

unless you're in the top bracket (I'm sure as hell not but for those who are it must be a pain in the rear)

Clay Pigeon, on 26 February 2013 - 10:56 PM, said:

There's no such thing a cheese. There are effective builds, and ineffective builds, and some middle ground between them. You were in a SRM boat a1, you just got out played by our A1.

saying something doesn't exist doesn't make it go away, though I do dislike that they're called cheese builds, cheese hasn't done anything wrong to anyone, we should come up with a new name for "cheese" builds.

[valkyrie]

Clay Pigeon, on 26 February 2013 - 10:56 PM, said:

There's no such thing a cheese. There are effective builds, and ineffective builds, and some middle ground between them. You were in a SRM boat a1, you just got out played by our A1.

Kong v. Kong is an entirely different thing. Maybe if Vass stopped walking into blatant killzones once in a while our team would've done ok. That's literally the only series of matches I've brought out any sort of OP build for in about two weeks.

[KinLuu]

This is how elo works.

If you want to use a bad mech, you will reach your elo-ceiling sooner than if you use a viable mech. And people with less skill then you will be able to reach the same elo ceiling than you, if they use a superiour build. Elo is the combination of skill and build.

If you want to push your elo even further, you need to use a viable build. If you are fine with your elo, you can continue to use your funbuild. But do not expect to win more than 50% of your games.

[Critical Fumble]

KinLuu, on 26 February 2013 - 11:06 PM, said:

This is how elo works.

If you want to use a bad mech, you will reach your elo-ceiling sooner than if you use a viable mech. And people with less skill then you will be able to reach the same elo ceiling than you, if they use a superiour build. Elo is the combination of skill and build.

If you want to push your elo even further, you need to use a viable build. If you are fine with your elo, you can continue to use your funbuild. But do not expect to win more than 50% of your games.

Even then, if its doing what its supposed to, you'll wind up plateauing at 50/50 again when you're matched against peers in other chee- er, competitive builds. edit: supposing you're not one of the absolute top players, in which case you'd hit the ACTUAL ceiling and win more often than not.

So, really, its a question of, "Just how competitive do I want to be?".

Edited by Critical Fumble, 26 February 2013 - 11:15 PM.

[valkyrie]

KinLuu, on 26 February 2013 - 11:06 PM, said:

This is how elo works.

If you want to use a bad mech, you will reach your elo-ceiling sooner than if you use a viable mech. And people with less skill then you will be able to reach the same elo ceiling than you, if they use a superiour build. Elo is the combination of skill and build.

If you want to push your elo even further, you need to use a viable build. If you are fine with your elo, you can continue to use your funbuild. But do not expect to win more than 50% of your games.

Except ELO doesn't take into account different builds, as has been said, just previous success. So, the moment I step out of a -3L or a Splatcat and go back to running my Cent, my ELO doesn't change. It just assumes "well he's still going to do just fine" and match me up against someone who is either way above my "actual" skill level or equally skilled and has no qualms about using such a build. I then get my face fed to a belt sander by a combination of min-maxers and the pubbies I'm still expected to carry.

You also mention how if I want to push my ELO further, I have to use a "viable" build. This is, beyond my personal gripes about how I'm unfairly manhandled for removing my cheese handicap, the ultimate issue. I'm fine sitting at a "comfortable" level when I can dart around in a Cent -AL and get into fun duels with other Mediums. The problem is those Mediums are going to become more and more rare. Why? Because the modern gamer ALWAYS wants to push their "rank" higher. To do this, they're going to run cheese builds early and often to get any edge they can. Who suffers for it? People like myself who refuse to bend balance mechanics to their very breaking point, i.e., most new players and people who don't care about stupid MLG type competitions.

You also risk homogenization of 'Mech variety - when a handful of builds do everything better than anything else, why would you ever play anything else? One of the most dynamic MechLabs in the series is effectively rendered moot in a heartbeat, because you have to play flavor of the month to win regularly. People do this in TT with custom builds too. Know what happens to them? People stop playing with them because it's no longer fun. We unfortunately do not have that luxury.

Telling people they have to play "standardized" cheese builds to have a hope of winning regularly is completely wrong on every level, and we all know it.

Edited by valkyrie, 26 February 2013 - 11:22 PM.

[KinLuu]

In every elo based game you can only play the stuff that is viable in the current meta, once you reach a certain elo level.

If you play non-viable stuff, you drop in elo - as intended. Elo is a measurement of competitiveness.

[Glythe]

valkyrie, on 26 February 2013 - 08:17 PM, said:

ELO (to my understanding) matches based on skill and balances teams accordingly. That means I could drop in a game with a moderately good ELO, but be grouped up with a bunch of newbies who have no experience or awful builds, and the other team gets "ok" players who have all just discovered D-DCs, Raven -3Ls, and Jumpphracts. Needless to say, we get slaughtered.

PGI really needs to re-evaluate the use of ELO in MWO. The more I play now, the more I realize it's doing far more harm than good. Right now, it's just going to further homogenize 'Mech choice (especially in PUGs), and we all know stagnation is the quickest way to kill a game.

I was under the impression that the MM matched total Elo score for both teams....and did not match up players by their number. This makes sense in that if players with high Elo team up you get worse and worse allies. This is a pathetic system as 1300 players should not be thrown on my team to balance my high Elo. The game should find someone with close enough scores for each player in the game with a 50-100 point variance.



But I agree Elo is hurting the game right now.

[Edustaja]

Yeah. Elo works as an aggregate for both your skill as a pilot and mechbuilder.
I've been having much fun dropping solo. I mostly get the non cheese opponents in decent mechs. Being overly competitive in 4-mans is just shooting yourself in the leg. I can do that in the 8-man queue.
My founder's mechs have actually seen a lot of play after the Elo hit. I usually only bring the D-DC, 3L, 3D or A1 if there is a real need for it.

[Jonathan Paine]

I don't know how to write this post without offending the op, but I will try.

You complain that you cannot play your non-dysfunctional cheese mechs, and still win consistently and increase your Elo. You complain that dropping with your team of well above average players increase your Elo rating. What do you want from the game? Unless you are the superawesomeelitest of all mechwarriors, you will lose as much as you win. Only those who are the natural grandmasters will win more than they lose.... as long as they drop in a good group.

Get used to winning and losing on an equal basis. Embrace a mediocre mech, and accept a lower rating. Embrace a cheese build, and enjoy dropping (mostly) against other cheesebuilds.

Funny things is, I was once of the same opinion. Much as I support the Elo system (having read up on it and gone: the system might never work in any given game, but it sure as heck will work wonders over 100+ matches), I worried about having to always bring my A-level game and my AAA+ 'Mechs.

And then I went "do I really want to spend eternity fighting splatcats, ecm ravens and ecm atlai?"

My answer was no, I'd rather have a lesser Elo rating and play whatever I want to play. So I'm levelling my Trebuchets, and suffer through some losses that might otherwise have been wins. If you want glory, wait for the next tournament - which I intend to sit out.

[Xelchon]

Guys I'm sorry to say ELO works like democracy. Not everyone can be happy.
But to be honest, look at the picture you have here. You said you have a wide variety of weapon systems.
Let's say the enemy has PPC stalker and Splatcat, one guy will go close combat on the PPC. The Splatcat is coming to defend him? Use the PPC guy as a meatshield against splats.
As you said, these guys are prone to panic and don't expect you to do such things. Use teamwork. PPC stalker will not know what to do, and won't be able to do anything even if he can, and while you distract them, your wingman will dps down the splatcat. The splatcat saw your wingman? He has two choices: Follow him or stay on you.
Choice one: Follow your wingman. Your wingman breaks off. PPC stalker is defenseless for at least a good enough period for the skilled player to wreck his day.
Choice two: He stays on you. You focus on evade. You can, if you're skilled, don't whine. Your wingman will weaken the cat easily. Everyone knows a cat doesn't soak damage well.
Something like this.
It's not possible? I did this with a pug guy who didn't even roll with me once before. It comes naturally since we both know our weapons and their combat potential.
Let's say if the enemy knows the Cat's and STK's weaknesses, and utilizes their forces accordingly;
They weren't at all no-skill players now were they?

Edited by Xelchon, 27 February 2013 - 01:19 AM.

[Steven Dixon]

Elo seems to work pretty well for pugs but not great for 4 mans. I think their are plenty of reasons for this, but I think a lot of it is inflated Elo's. If your W/L rate is above a 1 you have an inflated Elo. Even if your not W/L isn't above a 1 if your team members are above a 1 then your average is inflated (I'm not using the term in a derogatory way).

Their are a few ways to get an high Elo, one is to be abnormally skilled player, another is to group up with others (this is a significant advantage and distorts skill levels and inflates your Elo), the last is to use 'cheese' builds (this also offers an advantage and distorts skill level). If you are using one of the above methods you are likely going to be paired with other people using that fall in one of these categories.

[Red Line Pilot]

I don't think you understand how complicated it is to let a computer set up equally skilled teams of random players. ;-)

ELO assumes that you play each game as if it is your last. It doesn't differentiate between casual-saturday-evening-with-beer-and-non-al-lollybuilds drops and a tournament game. Or in other words, the intentions of a player arae a big factor in the balance that can not be measured by the (this) system.

For as far i know it also doesn't track an ELO-score for each combination of Pilot, Map, Mech and loud out (which would be the ideal situation) in this implementation of ELO.

And then, as what the OP is about, you have the problem of how do you compose two teams of random players. Things like ELO distribution, average ELO and, hack if you want to make it complicated, separate modifiers on ELO-values for the above factors and distribution parameters come to mind.

So, it is pretty complicated and almost impossible to make perfectly balanced teams. But it think the game is getting better with every patch and every refinement in the balance algorithm is something i welcome.

[Soulscour]

I play the game how I want to play it. Not with a handful of the most optimal builds. Anyone who thinks they got to stack the odds just to maximize thier rating is just going to be punishing themselves now. Top tier will probably be mostly cheesers. You guys have fun up there where your cheese build just fights other cheese builds. Splatcat vs Splatcat and stacked ecm ravens and AC20paults and all kinds of DDC atlas. I'll have plenty of fun right under that tier.

[Kristov Kerensky]

Optimized builds, which is what the so called 'cheese' builds are, will allow one to reach a higher Elo ranking than your actual skill would otherwise allow, but that's also as high as you'll get and it'll only hold true while you use those builds. Step out of them and you'll fall quicker then a feather on Saturn.

My nephew is a LoL fanatic, and we were actually discussing this issue specifically in relation to 'cheese' champions in LoL. He made a smurf account a few days ago and already has a higher ranking then his 'real' account that he's played for quite a while now. I pointed out that that was due to 2 factors..

1) He knows the game and the system inside out now, so his playing is better which is a small factor and..

2) The major factor is that he has used 2 specific champions known for being OP 'cheese' builds in LoL, champions he doesn't have on his 'real' account at all due to that very fact no less.

The ranking his smurf account has attained has very little to do with skill and a lot to do with using OP champions.

SAME thing is happening in MWO right now with the optimized builds, they are extremely forgiving of lack of skill, so using them gets one to a higher ranking then skill alone would rightfully earn, exactly as Elo based ranking systems work in every other online game that uses them. This isn't new, it's not a suddenly discovered issue with how Elo based rankings work, it's a known factor that's impossible to do anything about.

Remove customization and this situation wouldn't change one bit, there are optimal stock builds in each weight class and those would replace the optimized customs we see currently and would require no more skill to use, that is the point of optimal builds, skill isn't a major factor in their effectiveness.

If you have actual skill, then your rank will be high regardless of what you are using, minor variances will naturally occur and playing around with test builds won't really cause any major shift in ranking. If you don't have skill, you step out of those optimized builds and your ranking will drop, pure and simple.

The OP noticed that the people he usually drops with are getting beat a lot more then they used to. This means the ranking system is working as it should be, and that previously the OP and his friends were getting put up against extremely unskilled players far too often, which set them up for a higher then deserved rank. And they are seeing the results of that, more losses than they previously had...no offense, but it would seem your collective skill levels aren't that great.

I don't use the optimized builds, I use what tickles my fancy, which are usually quite un-optimized builds for the most part but they fit MY playing style and the tactics we tend to employ in Shadow Rat's Marauders. My own WLR has gone up since the introduction of the new Match Making system when dropping solo, and the WLRs of my brother Rats when we drop together as a premade have gone up as well. Yes we see lots of the optimized builds in our drops, but since we don't tend to use them ourselves, we aren't real afraid of them due to the fact that our skills are what got us to our ranks unlike the folks in those optimized builds...

[Renthrak]

Loxx, on 26 February 2013 - 08:24 PM, said:

Cheese builds are the easy solution for low to average players who have reached the top of their skill set. They can't survive in anything else, nor do they wish to spend the time getting better. So Instead they choose the easy route and go for cheese.

Losers in cheese builds isn't a problem, it's target practice.

It's skilled pilots who would be well above average in any 'Mech using a cheese build as god-mode that is the real problem.

[grayson marik]

Quite funny how - even casual 4 men - players would benefit if they could just meet up with other people in a room and agree to a "non cheese" match... or even a stock match with trials and have fun...

All I say to the cheese non cheese discussion is:

Remove teams from pugs and at the same time provide a lobby!

[icey]

its not even about cheese, its just the fact that right now there are about half a dozen 'strong' builds in the game, and if you bring anything that isnt one of those builds (say because you want to have fun and play something different), your whole team suffers.

it would not be difficult to temporarily modify your elo rating at the time of matchmaking by some number to account for how good/bad your chassis is. call it a chassis power rating or something - PGI knows how often each chassis wins or loses and how good they are relative to each other based on all their telemetry.

not to mention the obvious silliness of comparing a 0/8 basic new trebuchet to a 1/1 mastered 9A as though they were equal. (and they are, as far as your Elo can tell)

at this point, nobody knows how long it takes to adjust your Elo down by losing - it could take weeks for all we know.

in the meantime, you will be the guy everyone on your team groans at for bringing a dragon or a spider or whatever. so eventually you'll stop playing anything thats not a 3L, 9A, A1/K2, Jump 3D, PPC boat stalker or DDC. and it's stupid if the incentive system of the game encourages this.

8-man games are for 'competitive' play where everyone should bring their strongest mechs. If i want to mess around, i'll play 4-man or solo, but now you really can't. being forced to bring your A-game and your same few top tier mechs all the time is going to fatigue a lot of people pretty fast, and thats terrible for a game largely built around customising space robots.

thats the point valkyrie is making. It's pretty obvious and sensible.

Edited by icey, 27 February 2013 - 02:50 AM.

[Ghogiel]

It should come as no surprise since Cong members aren't quiet about this issue.

if you stomp to the higher Elo brackets, you have to be ready for what you meet there. And what you meet there are the higher level players, not only the players, but their tactics will be higher level. It's true many are still adjusting out of 4man pugstomper tactics, which does not work very well against higher level players in 4mans, all with complimentary min/maxed builds, and are going to be well drilled and oiled up ready for anyone trying casual play


Working as intended.

[J0anna]

valkyrie, on 26 February 2013 - 08:17 PM, said:

I think figured out what the single biggest issue with ELO is. Aside from the obvious tonnage matching issue, it's the fact that for mid to high ranked players, it's now become "cheese or die."

But now follow your thoughs through. The more you lose, the lower your ELO will go, while those in cheese builds will see their ELO shoot up. Overtime, you be facing them less and less, and end up on the team with 'ok' players. The problem isn't 'cheese builds' per say, it's that elo thinks you're better than you currently are (at least for the mech you're using).

No matter what PGI does, there will always be 'cheese builds' as people will gravitate towards what works well, and what doesn't. Get over complaining about them, because balance will never be perfect, something will always be better. I understand losing is no fun, but winning all the time isn't a good thing either, as people who always lose will leave the game and people that always win will become bored. Try looking at it this way, since you're with 4 new/poor players, you're going to have to elevate your game to carry them. It's a greater challenge to win when the team isn't helping as much.

Lastly, if you really find it hard to play/coordinate 4 PUGs, then stick with 8 man for the time being (this is the home of cheese builds), it's a pain atm, but it's all you have until ELO gets sorted.