60 replies to this topic

[Roland]

Honestly, the big problem is that instead of segregating off the lone wolves, they segregated the 8 mans, and they just flat out banned the 5-7 man groups.

The current system essentially creates additional incentive to sync drop.

If they segregated the lone woves instead, then the folks who really have no idea at all how to play can play with each other and not get stomped quite so bad (although McKenna is right in that folks are gonna ***** and moan anyway, cause bad players will lose no matter what you do).

Then you just throw all the grouped players together. The 8 mans will have more folks to fight against, and groups with 5-7 will actually be able to play together instead of the cluster that takes place now.

There will be some small groups that cry because they have to play against more organized folks.. but really, tough noogies for them. They get to see what it's like on the receiving end of that kind of fight.

[TruePoindexter]

Maybe when the game population gets larger they can simply make it so that teams only can contain a single premade. In a worst case that would lead to 4 premade 4 pug vs 4 premade 4 pug games which I think is fine.

[Roland]

TruePoindexter, on 20 May 2013 - 11:31 AM, said:

Maybe when the game population gets larger they can simply make it so that teams only can contain a single premade. In a worst case that would lead to 4 premade 4 pug vs 4 premade 4 pug games which I think is fine.

No, see, that's not really good enough.

If folks have 5 players, then they should be allowed to play as a group. Otherwise, you are basically hampering growth of the community.

Regardless, it's moot, since apparently Ekman already said they're bringing back groups with more than 4 players, since the matchmaker is better able to balance teams.

[Oderint dum Metuant]

Roland, on 20 May 2013 - 11:36 AM, said:

No, see, that's not really good enough.

If folks have 5 players, then they should be allowed to play as a group. Otherwise, you are basically hampering growth of the community.

Regardless, it's moot, since apparently Ekman already said they're bringing back groups with more than 4 players, since the matchmaker is better able to balance teams.

True, and the sooner any group of any size can play against other groups, and not random lone wolves the better.

I can't even fathom why they segregated 8 mans, rather than the single players..

Edited by DV McKenna, 20 May 2013 - 11:48 AM.

[Kraven Kor]

Joseph Mallan, on 20 May 2013 - 05:49 AM, said:

Tuesday nigh Marik drops still happening?

The "Marik Madness" drops are on Monday.

On Tuesday, we conspire to see how many people we can force to rage quit.

(Not really.)

[TruePoindexter]

Roland, on 20 May 2013 - 11:36 AM, said:

No, see, that's not really good enough.

If folks have 5 players, then they should be allowed to play as a group. Otherwise, you are basically hampering growth of the community.

Regardless, it's moot, since apparently Ekman already said they're bringing back groups with more than 4 players, since the matchmaker is better able to balance teams.

Not sure I agree with that - 4 coordinate good players are better than 4 individual good players. This difference becomes even more dramatic the larger the teams get. This seems like a step backwards rather than forwards.

I do think though that the solo queue would help a lot too. I would argue that the game could stretch it further and follow the LoL model having a ranked solo queue for competitive players and an unranked solo queue for casuals. Assuming the game population is large enough to support this.

[Nubsternator]

hammerreborn, on 20 May 2013 - 06:37 AM, said:

Prior to the most recent MM change, in 150 recorded matches, only 8 were above 100 tons, with the maxium being 135 ton difference between the teams. Most notably though, is that in each of those 100+ ton games, the heavy team lost. I refuse to believe that a 300 ton MM imbalance exists outside of 8v8s, which is an entirely different ruleset. So no, you don't see a 300 ton MM imbalance either.

See, facts and data, I has them, the average difference in teams overall was 28.9 tons with the heavier team winning rougly 51.5% of the time

It should also be noted that 0 ton difference games were more prevalant than 100+ games.
|
|
|
v

I too doubt about 300 ton mismatches happening as often as people say. I can see Awesomes put against Atlases, as they are the same tier of mechs. You get a lot of that in a single match and it adds up.

Over half of all statistics are made up on the spot. Saying you have sources and showing these sources give two different responses from me.
I don't doubt what you said, I just feel better when we too can check these stats out, as a simple search online yields nothing useful. You also say "recorded matches" and don't give any details as to which matches. Were they in a row? Random matches? Whatever you felt like at the time?

[jeffsw6]

Kraven Kor, on 20 May 2013 - 11:53 AM, said:

On Tuesday, we conspire to see how many people we can force to rage quit.

Now we know who took our HUD.

Nubsternator, on 20 May 2013 - 05:53 PM, said:

I too doubt about 300 ton mismatches happening as often as people say. I can see Awesomes put against Atlases, as they are the same tier of mechs. You get a lot of that in a single match and it adds up.

I think it happens because of 4-mans with a lot of lights just happening to be on one team, and none of that on the other side.

I have been playing in Conquest mode for the past couple of days, and match after match, I'll end up with or against a 4-man of three or four light mechs. Even on big maps, they seem to get wrecked, because noobsauce light mech pilots want to brawl and die. On small maps, game over.

[hammerreborn]

Nubsternator, on 20 May 2013 - 05:53 PM, said:

I too doubt about 300 ton mismatches happening as often as people say. I can see Awesomes put against Atlases, as they are the same tier of mechs. You get a lot of that in a single match and it adds up.

Over half of all statistics are made up on the spot. Saying you have sources and showing these sources give two different responses from me.
I don't doubt what you said, I just feel better when we too can check these stats out, as a simple search online yields nothing useful. You also say "recorded matches" and don't give any details as to which matches. Were they in a row? Random matches? Whatever you felt like at the time?

Did you look at where the arrow is pointing (see my sig). All results are given and the initial screenshots are all linked too.

And even if every atlas was paired against an awesome, thats a 160 difference, tops.

For a 300 you would need half the team to be jenners against all atlas.

Edited by hammerreborn, 20 May 2013 - 07:24 PM.

[w0rm]

I just troll people in allchat when I pug with "wow. cant believe the syncdrop worked again"

[MrZakalwe]

Wonder why I've only seen a couple of these fabled synch-drops in the hundreds of games I've clocked up in the last month and a half.

Matchmaking must like me, I guess.

[hammerreborn]

MrZakalwe, on 21 May 2013 - 04:53 AM, said:

Wonder why I've only seen a couple of these fabled synch-drops in the hundreds of games I've clocked up in the last month and a half.

Matchmaking must like me, I guess.

Many people call out sync drops when they don't exist. 7 davions on a team, SYNC DROP!

Because its completely impossible for two groups with one of the most popular factions in the game to be matched on the same team.

The only time I suspect syncs are when its 8 liao yelling squawk

[cyberFluke]

Kommisar, on 20 May 2013 - 05:58 AM, said:

You do realize that ELO is still affecting the match making. If a group of players does sync-drop, the vast majority of groups doing it are comprised of largely veteran players. If all 8 of them do happen to end up on the same team, they are not going to be pitted against 8 solo guys still grinding through their trial mechs. The exception is if this all happens during the really, really low player periods; I'm guessing 2 am EST or something, maybe? Not sure what player numbers are like across various time periods.

Otherwise, while the ELO system is not perfect, it's not going to match an 8 man vet sync drop against even 3 guys in trial mechs unless there are something like less than 50 people online playing.


I'll also add that the accusations of 8-man syncs are far more prevalent than the reality. I've been accused of this a LOT by being on the winning team. Sometimes, you just flat out get rolled and it doesn't take a sync drop to happen.

You do also realise that if say, three vets invite the FNG to their team who been solo dropping and getting stomped, the team average ELO is vastly lower than it should be, so gets dropped against "lower skilled" players?
So two teams (of two vets and two FNGs) sync dropped with a third party doing the countdown (to avoid lag causing timing delay) will actually frequently meet, and fight lower ELO players than the vets should be. Not that I know how a lot of these "High Skill" Merc Corps are padding their stats already or anything....

[MrZakalwe]

hammerreborn, on 21 May 2013 - 04:57 AM, said:

Many people call out sync drops when they don't exist. 7 davions on a team, SYNC DROP!

Because its completely impossible for two groups with one of the most popular factions in the game to be matched on the same team.

The only time I suspect syncs are when its 8 liao yelling squawk

If it helps I think I've seen a failed synch drop with 4 liao on each team yelling squawk followed by a little swearing.

They played against each other so didn't make any difference to my day.

[hammerreborn]

MrZakalwe, on 21 May 2013 - 05:16 AM, said:

If it helps I think I've seen a failed synch drop with 4 liao on each team yelling squawk followed by a little swearing.

They played against each other so didn't make any difference to my day.

I mean, that can happen just by chance.

The goons are plentiful AFAIK, if there are lets say 40 on at any given moment, if they don't segregate themselves then it's likely they all share a similar Elo from playing amongst themselves and sharing in the wins and losses. It's not entirely unreasonable that even just randomly dropping over the course of the night one of those teams is going to find another.

I've had plenty of "sync" drops with the Robinson Rangers just while playing by myself. I play with a certain 4-5 people all the time so they're all in my bracket anyways.

I'm not saying it doesn't happen, but its not the end of the world as people like the OP seem to portray, and certainly never resulting in 300 ton swing matches like one person suggested.

Edited by hammerreborn, 21 May 2013 - 05:22 AM.

[Taemien]

I doubt sync dropping is happening as often as players say it does. I know House Marik likes to do sync drops of solo players for their events. They don't care which team they pop in on though. Thats the closest thing to a sync drop.

When 12v12 comes out you won't have to worry about two 4 mans sync dropping. They'll be allowed to drop as 8mans. But it won't be doom and gloom. Players who drop in organized teams will be higher ELO so those of us that don't won't fight them. If we do, then it just means we're that damn good and can deal with them.

Seriously, teams, even VOIP organized ones can be beaten. A buddy of mine and I were running duo and ran into a 4 man lance in River City once. I could tell they were a premade, all Dracs, similar names, running together in a formed up lance. They did not see our Raven and Centurion coming. We tore up their Treb quickly and then their other mechs before their DDC could get in there to support (we had hit them as they left their base).

Good thing too, since the rest of our team were PUGs. They'd been demolished by these guys if they got dugin to where they were going. Point is, a team can be bested by smaller premades if you hit them fast enough. And judging by the lance setup, there's lots of small premades out there, and I mean Alot.

When you Solo Queue, you're not really solo. Except maybe in the lower ELO brackets.

I almost wonder if its matching us up by CW rules. Several times I'll see a Lance of us 2 Davions, then 3 Davions, then 3 Stieners sometimes. Pretty sure its coincidental but pretty cool nonetheless. But the point here is, you usually have a premade or two on yourside. Find out who it is and see what their doing. They're on voice so are going to be very coordinated, support them and you'll likely win.

When fighting a 2 lance premade (8 man sync drop), a good tactic is to hit them hard and fast. 8 people means someone is probably micro managing. Thats alot of players to manage and get setup and you could easily hit them before their ready. Though if they are working in autonomous lances, then you have an issue, but then its just like playing against two normal premades which you should be used to anyway. There's usually 2-3 premades per teams in 2s, 3s, and 4s.

[silentD11]

hammerreborn, on 21 May 2013 - 04:57 AM, said:

Many people call out sync drops when they don't exist. 7 davions on a team, SYNC DROP!

Because its completely impossible for two groups with one of the most popular factions in the game to be matched on the same team.

The only time I suspect syncs are when its 8 liao yelling squawk

I know of two major Liao groups one of which is the goons, both of which have public VOIP options stated in the forums. Both of these groups can have actual 8 mans going on along with multiple four mans. I also know of a few smaller Liao groups that function off of other VOIP options.

So it's completely possible to have 8 Liaos on one team and have it be from 1 group by accident, or even from three different groups that aren't related or on the same voice. I've also seen 8 on one side and four on the other, and at times the 8 on one side aren't from the same group or on the same VOIP either.

And I'll be honest, I know a lot of us like dropping against or Liao's to grudge match the other groups or fight our own people. In fact we generally prefer that to playing on the same side.

[Vodrin Thales]

TruePoindexter, on 20 May 2013 - 11:31 AM, said:

Maybe when the game population gets larger they can simply make it so that teams only can contain a single premade. In a worst case that would lead to 4 premade 4 pug vs 4 premade 4 pug games which I think is fine.

This is largely what happens now. 8 man synchs in the 4 man queue are currently rare as the matchmaker will frequently separate them even when they are trying to drop together. Only times it will happen frequently are low pop times. It's not the problem it used to be.

8 man synchs are a bit cheesy and lame, as if you have 8 drop in an 8 man. If you don't your simply acknowledging you are second rate players.

As for tonnage mismatches I have seen some really large ones during off hours, but not much during primetime in the US.

[Chavette]

buttmonkey, on 20 May 2013 - 03:58 AM, said:

come on pgi please sort this out for the love of kerensky!

how is it that you can acknowledge that an 8 man team shouldnt be dropping with pugs or 4 mans by keeping them in their own seperate queue, but you continue to let the same 8 man groups do sync drops into a match full of pugs. it beggars belief.

nothing says "**** off" to new players more than being shoved into their first match against a enemy comprised of 4 man premades and 8 man sync droppers.

its so damn easy to fix! you have 8v8 ok fine you got that.

then you have 4man groups dropping against other 4 man groups only.

then you have 2man and 3 man groups dropping against other 2 and 3 man groups.

then you have puggers dropping against other puggers only.

i really hope im not alone in thinking that this is a very easy solution to a very big problem.
its bad enough that there is no tutorial but to be thrown in match after match against an organised group must be a complete nightmare for new players.

This is the exact same thing I've been sayin for half a year, it didnt get to their heads yet.

Its in my sig, named group matching.

[Kommisar]

And that is the kicker about any match making system that people tend to forget. A match making algorithm (not just MWO) will always be limited by the seeds (players) in how well it can generate matches. During low-player hours, the match making is going to be worse. It has to be. The alternative is to simply not allow matches until more people are online and dropping. Which, I might add, would cause a huge volume of QQ threads in these parts.

If there are only 36 people actively dropping within the 5 minute window (or whatever it is set to now), the chances of forming 4 good matches, with ELO and tonnage within acceptable margins are just flat out lower. I don't care if Jesus, Godel, Turing, and Ryan Gosling got together and created the single greatest mathematical system for match making in history, winning a Nobel Prize in the process, it would still create poor matches when there are low seeds to the system.

As for why they segregated 8 mans from the solos; it had everything to do with numbers. The 8 mans were a much smaller portion of the player population. Plus, on the math side, if you have 7 to 8 of the players on one side "locked-in", the match making algorithm can only work on one side of the match. Throw in, say, a 3 man group on the other side and now it only has 5 slots to fill to create a good match. As you constrict the match maker like this, the odds of creating a good match drop. And it drops non-linearly. Till you get to 8 man versus 8 man; no match making is involved and the odds of a good match are completely random. If you throw in a 6 man team versus a 7 man team, the odds of a good match are so low that, for all practical purposes, it is also random.

Finally, you have the non-mathematical elements. No algorithm is going to be good enough to be able to match a 7 or 8 man group against a fill in team of solos to 4 man groups with any consistency. Even with a huge seed pool to draw from. The inherent advantages of being a cohesive group with VOIP are just too great in the average case.

Please note, I never once said "impossible" anywhere in my post. With any system there are always outliers. And, as we are human, we tend to remember the extreme instances the most. We also tend to fill in unknowns with presumptions as our minds are hardwired to create order from chaos. This is why every culture and society on the planet has created their own vast history of superstitions. They are constructs of our minds to force a logical and orderly facade over otherwise completely random events.

Take the 300 ton imbalance match. There may have been one. There may have been several. And, as an extreme event, those who were there are going to remember them in such a way that they have more "weight" in their over all perceptions than the thousands of other matches they have played. The same way that you will always remember your drive into work the day that you hit a deer with your car. You have made that trip hundreds of time and the deer strike represents less than 1% of those days. Yet, if anyone ever asked you about your commute, the deer strike will immediately come to the front of your thoughts. The other 99% or more of the commutes, you mind has literally forgotten by the time you go to lunch. I remember a study where they asked a large pool of commuters how many, on average, red lights they hit on their commute into work. Almost everyone responded with a number well over half.

So, back to match making. What I read here is that everyone wants what they want when they want it. If they drop solo, they want to drop into a match full of other solo players. If they drop in a group, they want to drop against... well, an "equal" force. They want a computer algorithm to be able to tell how good an given player is, how good that player is when playing with a group of other players, then, inside about 1 minute, create a fair match for them to play in with a 90% plus success rate. Also, the general definition of a fair and good match (if we are being honest with ourselves) is one in which we win. We are, after all, humans. There will be a percentage of losses that we, as a group, will assign as "fair" even when we lose. But the chances of judging a match as "fair" go up significantly in a win. I don't have any studies to back that assertion up; but if we are all honest with ourselves, I think we can all agree that as a whole population, this isn't far off the mark.

My advice here is to lower your expectations. If you are holding out for perfection in a system populated by humans that can all make individual choices without a culling mechanism to remove disruptive (define this however you like) elements, then you are going to be waiting a LONG, LONG time.