536 replies to this topic

[AlmightyAeng]

Flaming oblivion, on 24 April 2014 - 04:47 PM, said:

There is for me if I play at the wrong time all the pre mades turn I know most of them who they are learnt fast. Couldn't understand initially why my results were so different when facing them and at that time of day till I found out.

I played 8-9 games last night on the public test server. In 5 of them, there were no premades. People had trouble getting in game in a timely fashion, so they started dropping solo. Those 5 games were ALL stomps. I saw names I recognized, and we talked about how we were dropping solo.

Maybe the people who group up regularly and practice with their buddies are simply better than you, regardless of whether or not they're grouped...and before the more serious players come online, you're a big fish in a small pond.

Just a thought.

[Magna Canus]

Jun Watarase, on 23 April 2014 - 07:54 PM, said:

At this point i hope nobody still believes that premades are a myth when some players are very well known to only drop in meta builds and stomp newbies, you know the ones in 4 man poptarts with PPCs/AC5s....cancer killing MWO.

I know the kind of 4-mans you are talking about. The "GGclose" types. Had a couple of games with and against them (PUG of course) this morning. In the game I played on their team in my Misery the "Lord blah di blah, GGClose" turned up in chat and everybody knew what was up. We were down 6 (All charlie lance plus half of bravo) to 2 (half charlie lance) kills at one point and a cocky somebody on the other team typed in "GGClose B*tches" and everything seemed to kick into high gear after that. I took a center position with those guys on the fringes and a can of whop a$$ got opened up. The game ended with 1 of them and me left (PUGing as noted above), me with 5 kills under my belt *glow*. The enemy team had several LRMs and a few pop tarts to boot, so mostly "meta".
Later on I had 2 matches vs. them, 1 on HPG and another on alpine. On HPG we all went to the rat cellar and got creamed royally, but I still managed to kill a battlemaster. Alpine was just a disaster. LOL

I agree that these guys are going to ruin a new guys game pretty fast. For some players PUGing with/vs them is a serious and desired challenge. I get my face eaten too, mainly because I was in the wrong spot at the wrong time, got separated from my lance/group, etc. Suck it up and soldier on dude. Whack-a-mole and LRMs beat pop-tarts every time.

What gets my goat is all the "e-racism" going on here vs team players. There are meta-humpers out there pop-tarting and LRM spamming the life out of the enemy, but you had better be honest with yourself and realize that solo players do not magically not run these builds just by virtue of dropping solo. On the contrary, many of them run these builds for that self same selfish reason people are accusing groups of; win at any cost, maybe justified with "they have the teamwork advantage". Dropping solo you don't have to worry about your buddies asking you "aren't you tired of that build? Don't you think it's a bit unfair"? I am in a unit and have yet to own a highlander/victor/CTF-3D and it is exceedingly rare to see those mechs in any of my units drops, especially because they are generally frowned on. Factual proof that there is at least one unit out there that does not fit the "GGClose" meta-hump myth.

When you are really concerned for NEW PLAYERS have a look at this.
http://mwomercs.com/...cadet-training/

[Joseph Mallan]

Ghost Badger, on 25 April 2014 - 05:09 AM, said:

I played 8-9 games last night on the public test server. In 5 of them, there were no premades. People had trouble getting in game in a timely fashion, so they started dropping solo. Those 5 games were ALL stomps. I saw names I recognized, and we talked about how we were dropping solo.

Maybe the people who group up regularly and practice with their buddies are simply better than you, regardless of whether or not they're grouped...and before the more serious players come online, you're a big fish in a small pond.

Just a thought.

Me too. I was pleasantly suprised how rapid the games ended! In 30 minutes I had almost 5 games in. 2 ROFLStomps 3 close maulings! All in all 5 good matches!

[AlmightyAeng]

Joseph Mallan, on 25 April 2014 - 05:41 AM, said:

Me too. I was pleasantly suprised how rapid the games ended! In 30 minutes I had almost 5 games in. 2 ROFLStomps 3 close maulings! All in all 5 good matches!

Yeah...I got mostly conquest matches, lol, and all everyone wanted to do was kill things to test how 'fair' 3/3/3/3 was and find out if SRM's were worth a damn, lol. Took the splatcat out 4 times...330, 407, 470 and 5-something...with a smattering of kills and assists.

Nothing spectacular, but not ENTIRELY useless.

The 3 SRM4's on the DDC were noticeably less crap...but, being buttressed by the other weapons that actually damage what they hit, the numbers were, of course, higher.

Edited by Ghost Badger, 25 April 2014 - 05:45 AM.

[PropagandaWar]

Jun Watarase, on 25 April 2014 - 02:05 AM, said:

I think this picture says it all really....and the worst part is that half our team was stuck in a pass for most of the game, nowhere near any of the cap points, and we still completely stomped them with no effort...

Why you hiding names. This is just a silly assumption that one team has a premade on it. Guesstimation at its finest. People are great at sucking, as well as they are great at having other people be the fault of their suck.

[Flaming oblivion]

Ghost Badger, on 25 April 2014 - 05:09 AM, said:

I played 8-9 games last night on the public test server. In 5 of them, there were no premades. People had trouble getting in game in a timely fashion, so they started dropping solo. Those 5 games were ALL stomps. I saw names I recognized, and we talked about how we were dropping solo.

Maybe the people who group up regularly and practice with their buddies are simply better than you, regardless of whether or not they're grouped...and before the more serious players come online, you're a big fish in a small pond.

Just a thought.

Perhaps your right perhaps its a simple of case of in the day I get fair fights where everyone competes pretty evenly and fights can go either way, Once it gets to night all the pug's in my bracket log out because they know the pre made "boogie man" is arriving on line making there next few hours of battling a stream of epic fail stomps. Now why should PGI change that , Its completely reasonable that for a certain proportion of the day if your PuGing in a certain ELO bracket you have to die . (end sarcasm)

[Graugger]

WHY I CANT GET ANYONE TO PLAY MWO FOR LONG

Would you like to try and play a piano with no keys?

[Hordamer Mendelbaum]

I really like the thoughts that the OP put out there. It's a good statement of many of the negative impressions of the game.

However, I don't think the problem is really about the synch-dropping pre-mading meta-builders as much as it is about the salient point that, under the current matchmaker, we have no way to opt out of the meta game. Said another way, we have no expectation of a reasonable match, playing the way that we believe is fun.

But the problem with the game is deeper than that. If we all just played the way we wanted to play and the matchmaker were perfect, then we'd be matched up with others that have a similar skill level and play style. However, the matchmaker is fatally flawed, for several reasons:

First, the matchmaker uses Elo as the matching metric. Elo is a numerical indicator, calculated from a win-loss history, which predicts the probability of the outcome when two players of a single player game are matched up. Since MWO is a team game, Elo is a questionable metric (a more appropriate, but impractical, metric would be to assign an Elo to each permutation of players that are grouped together). But, if we assume that the result of each match is due to some combination of team and individual play, then the resulting outcome can be (weakly) predicted based on the net contribution of individual skill levels on a team. As a result, the Elo metric in use will--eventually--converge to the correct value predictive of a player's individual skill. However, this convergence will be slow and noisy due to the weak dependence of match result on individual skill. That said, the matchmaker's performance will always be limited by the fact that it cannot account for team play, due to its sole use of an individual statistic.

Second, the matchmaker uses average Elo when making its determination that teams are balanced, but there is no a priori indication that average Elo is a valid predictor of match success at all. Specifically, the balancing metric should be able to predict the probability of the match outcome before the match even starts. If average Elo is a good metric, then two teams with exactly the same average Elo have exactly equal probabilities of winning the match (i.e. 50/50). This prediction can be validated relative to the outcomes of live matches. However, there has been no demonstration by the development team (who are the only ones who have access to the validation) that the balancing metric has any quality whatsoever.

Third, the determination of balance does not appear to have any consideration of the Elo variance across the two teams. Formerly, the match maker directly restricted the range of Elos (i.e. individual player skill levels) that would be present in any match making attempt. Recently, this restriction was relaxed due to wait times for the top tier players. (Naturally, the best players in the game represent the smallest population of players; 24 of them to showing up at precisely the same time during off hours is a low probability occurrence.) As a result, it is perfectly feasible to match great players with inexperienced ones.

Finally, the combination of using a metric that doesn't include team play, using average as a matching metric Elo, and not limiting the variance of individual skill levels is most likely to negatively affect the user experience of two groups of players: the worst (newest) players and the best players. Given that most teams will be made up of predominantly average players, if a new player tries to join the game, he will be much more likely to find a match when grouped with a top player. As a result, the new players are tossed in to the game playing with one or two top players on their team, facing a group of primarily average players. The result is a destruction in the quality of the match. New players do not fare well against average players, and top players have no chance when massively outnumbered.

All of these conclusions come from knowledge of the design of the matchmaker and a an understanding of population statistics. Design changes such as switching from 8v8 to 12v12 and increasing the allowable Elo spread both had the result of increasing match variance and thus decreasing the quality of the match making. Furthermore, segmenting the community into smaller populations will further increase the match variance. Mark my words: although it won't be as scary as the 12-man Queue (which is drawing from a tiny population), the Group Queue will be a mess.

TL;DR - The matchmaker is poorly designed. There's no point in playing in anything but pre-arranged matches until it's fixed.

[Lightfoot]

I know one MW 4 player who didn't like the initial difficulty and left, but I expected it. I think that you have to start out in MWO in a good Medium as your first owned and customized mech (unless you save a lot of CBills or buy a mech) RP's well with MechWarrior singleplayer games where you always start out in a smallish mech.

The balance of the mechs is very good in MWO where a Light can beat an Assault, etc.

The inherent difficulty in MechWarrior is understanding that good piloting keeps your mech upright a lot longer, and this is key to enjoying the match obviously. Good piloting is:

1. Always attack on a lateral tangent so your mech is moving across your opponent's HUD and not straight at them. This spreads the damage of different weapons fired in a group and forces your opponent to execute accurate shots instead of just laying the reticle across your mech's CT.

2. A Joystick makes you a better pilot faster and is the natural controller for a Mech. It can be setup to be nearly as accurate as a Mouse with a little adjusting, but is slower aiming. The choice is yours, Mouse is fine if you like that, just don't forget to move laterally on attacks which is too common with a Mouse.

CW will hold more players and get more players online regularly.

MechWarrior is pretty easy to play, easier with a joystick in my opinion, but not everyone will like it. I like it for it's sci-fi combat and figure this is the kind of game a PC was meant to bring to life. What would make it more immersive was if my pilot ejected from the mech on destruction and I got a brief aerial view of the battle before switching to observer. That and the occasional core breach!

Edited by Lightfoot, 25 April 2014 - 08:43 AM.

[Burakumin1979]

Hordamer Mendelbaum, on 25 April 2014 - 08:40 AM, said:

Second, the matchmaker uses average Elo when making its determination that teams are balanced, but there is no a priori indication that average Elo is a valid predictor of match success at all. Specifically, the balancing metric should be able to ...

Hordie. I agree with you 100%..BUT..

I promise you 99% of the people reading this think that " a priori" is a typo.

[The Trumpet of Gabriel]

I pug about 98% of the time. When I first started back in open beta I was not able to do much, but I knew that there would be a fairly steep learning curve going in. I watched some youtube vids of experienced players and started to learn what to do and what not to do and gradually was able to start contributing more to the games I was in. That said, I still have bad days and good days, but in my experience they usually even out in the long run. The best lesson I've learned as a pugger is that if there is a premade group on your team that knows what they're doing (some don't) stick with them, shoot at what they do, and don't wander off by yourself. My best performing games have all been as a solo pilot and not in a meta build. Just my 2 cents.

[Votanin FleshRender]

I honestly don't think premades are much of a problem... the far bigger problems, IMHO

1) new players start in the middle of the bracket, and have to endure getting schitstomped until their Eli drops.
2) related to 1, new players drop in the same games as vets. Yes, my side will have some vets too, but the vets on both sides can easily capitalize on the numerous mistakes that new players like me make, making for a quick death for new players.
3) no ingame VoIP.

I'm not sure I should still call myself new, but I do because I don't get to play much and my experience is still somewhat limited.

[Burakumin1979]

I'm actually looking forward to leagues forming using the private queue that develop rules for mech loadouts (limiting meta builds).

PGI's initial mistake of allowing complete customization of Inner Spehere mechs without limiting hardpoints within types has really led to this disaster of AC40 Jagers, 4-6 PPC mechs, and the newly revived Dual Gauss issue.

Had they broken down hardpoints into class 1-2-3 within the 3 main types, they would have been able to control the meta and balance more effectively.

Example:

Jager has six B2 Hardpoints across its arms.
A B2 point can fit B2 and B1 weapons.
B1 weapons are MG and AC2. B2 are AC5/UAC5 and AC10/LBX and B3 are AC20 and Gauss.

This allows for customization while still retaining control over mounting canonically ridiculous builds.

PGI has a ton of effective balance measures at its disposal but has decided to break them form the beginning.

Another example is the odd implementation of the heat scale. Ghost heat doesn't even have to exist if the heat scale isn't raised from 30 (for no reason at all) and dissipation works correctly.

Ballistics could be balanced by adding a realistic chance of ammo explosion/gauss rifle on crit OR reducing ammunition per ton to tabletop OR LESS.

EVERYTHING could be more easily balanced by adding rearm and repair and making the loss of a AC20, Gauss, ER PPC an expensive proposition.


There are tons of ideas, but with PGI not explaining, and more alarming if true, not possessing a clear vision of what they want the game and balance to look like, nothing will really be fixed.

[MischiefSC]

New players won't drop with vets after 4/29. The point of matching into tiers is 1000-1500 Elo drops in one pool, 1500 and up in another. Sub-1000 drops in a third. This gives a wide range of potentials for each drop but that 1500 mark is literally the indicator for 'above average'. This will literally mean that people in their cadet matches will never drop with veterans - unless it's really crappy ones.

Without grouping in a range in that 1500+ bucket you also avoid the 'Elo hell' of being just below that huge gap that will split top tier from upper tier players and serving, essentially, as food for top tier players. They'll get mixed in with everyone one who's 1500 or higher.

3/3/3/3 and 1 premade per team is going to largely gut syncdropping, especially if they're filling matches round-robin instead of FCFS (first come first served) which, with 3/3/3/3, would be more logical and result in total matches filled more quickly.

I'm eager to see how it works out.

[Hordamer Mendelbaum]

Burakumin1979, on 25 April 2014 - 09:24 AM, said:

...they would have been able to control the meta and balance more effectively.

While balance and a meta game that's not absurd would be really nice, having a balanced game is secondary, in my opinion, to not getting constantly stomped in unequal matches. As long as the match maker can properly account for the fact that some players reject the meta, or reject having to form groups, or haven't yet developed the skill or experience, and places them with and against players with approximately the same capabilities, then the game would still be fun for those players, even though they aren't playing 'optimally.'

Edited by Hordamer Mendelbaum, 25 April 2014 - 09:37 AM.

[Votanin FleshRender]

A new player question about R&R... didn't that lead to problems with players hanging back, not wanting to fight for fear of damage costs?

Or tons and tons of sniper builds? I can't imagine there was much brawling at all.

[Joseph Mallan]

Some players yes... but they are the same people who would stand behind "the guy" in real life. What broke the system was people just riding the 75% free R&R and not fixing the last 25%. It was dumb, but hey some folks will game anything they can.

[AlmightyAeng]

Votanin FleshRender, on 25 April 2014 - 09:51 AM, said:

A new player question about R&R... didn't that lead to problems with players hanging back, not wanting to fight for fear of damage costs?

Or tons and tons of sniper builds? I can't imagine there was much brawling at all.

Hanging back, no. Not refilling ammo or repairing fully, yes. There was a TON of brawling...a broken STD cost WAY less than a broken XL (which most snipers took due to weight and ammo constraints).

Edited by Ghost Badger, 25 April 2014 - 09:54 AM.

[MischiefSC]

Just popped in for a few matches.

Was reminded why I've been playing something else for a while.

Will be back after 4/29 with hopes that it'll be fun again.

I have nothing but the deepest empathy for new players who try to get into this game. Again, hoping the enhancements for 4/29 all come out, it'd be a good step in the right direction.

[Jun Watarase]

Lightfoot, on 25 April 2014 - 08:41 AM, said:

I know one MW 4 player who didn't like the initial difficulty and left, but I expected it. I think that you have to start out in MWO in a good Medium as your first owned and customized mech (unless you save a lot of CBills or buy a mech) RP's well with MechWarrior singleplayer games where you always start out in a smallish mech.

The balance of the mechs is very good in MWO where a Light can beat an Assault, etc.

The inherent difficulty in MechWarrior is understanding that good piloting keeps your mech upright a lot longer, and this is key to enjoying the match obviously. Good piloting is:

1. Always attack on a lateral tangent so your mech is moving across your opponent's HUD and not straight at them. This spreads the damage of different weapons fired in a group and forces your opponent to execute accurate shots instead of just laying the reticle across your mech's CT.

2. A Joystick makes you a better pilot faster and is the natural controller for a Mech. It can be setup to be nearly as accurate as a Mouse with a little adjusting, but is slower aiming. The choice is yours, Mouse is fine if you like that, just don't forget to move laterally on attacks which is too common with a Mouse.

CW will hold more players and get more players online regularly.

MechWarrior is pretty easy to play, easier with a joystick in my opinion, but not everyone will like it. I like it for it's sci-fi combat and figure this is the kind of game a PC was meant to bring to life. What would make it more immersive was if my pilot ejected from the mech on destruction and I got a brief aerial view of the battle before switching to observer. That and the occasional core breach!

Good piloting only increases the time it takes to kill you when you can move at high speeds (100+ kph) and are relatively small to hit. Imagine if 4 mechs fired 4 med lasers each and kept the crosshair on the CT of a light mech for the full burn time (not hard unless JJs are involved). That mech would go POOF.

Instead now what happens is that the light mech takes yellow damage all over its body, not just the CT and ends up tanking more firepower than an Atlas.

Ever tried to headshot a AFK mech? You would have noticed that the damage spreads all over the torso or mostly on the CT. Meanwhile drop into a training match and the AFK mechs there get headshotted easily.