Playing With 2 To 4 Friends Is Now Very Harsh
#41
Posted 03 July 2014 - 07:54 AM
#42
Posted 03 July 2014 - 07:55 AM
Ghost Badger, on 03 July 2014 - 07:26 AM, said:
Didn't I read somewhere that 12-mans would face other 12-mans as a priority, with smaller groups vs 12-mans as the 'release valve' ?
Example:
5x12 mans are in the queue: 4 of them are matched against each other, 1 of them will face a composition of smaller groups as release valve
12 man number 6 enters the queue, finds no other 12 man because 5x12 mans are allready in a match and thus will be matched against smaller groups.
Now let ELO kick in and the whole process gets even more complicated.
So depending on how fast the release valve opens (which I don't know) it is quite possible that more than one 12 man ends up facing smaller groups as long as the total number of 12 mans stays small.
More 12 mans = chances shrink that the release valve will open before they are matched against another 12 man.
Again (anecdotal) evidence: 6/7 drops the MM did put us up against smaller groups yesterday (20:00 GMT - 23:00 GMT)
Someone correct me if I'm wrong with that logic
#43
Posted 03 July 2014 - 07:55 AM
Bilbo, on 03 July 2014 - 07:52 AM, said:
Yeah, but won slightly more. Sorry, didn't mean by a HUGE margin. MM tries to line up 50/50 W/L based on ELO...and with skill/teamwork he manages to push it slightly into the win category.
Working as intended.
#44
Posted 03 July 2014 - 07:56 AM
KhujjoBD, on 03 July 2014 - 07:52 AM, said:
Happens. Earlier in the day (well before peak hours) we had a group that lost 10 of 11 in group queue.
Reality is, times of day REALLY impacts who you are likely to face. Peak hours are likely best, simply because the pool of casual/competitive types is much bigger, resulting in better Elo match ups. Play at a weak timeframe for the game, expect to see the same faces more than once and much broader Elo sweeps by the MM to make games happen. Unfortunately trade-off and one that is hard to avoid.
When faced with just indomitable odds periodically change the settings to just get "off-cycle" from a particular group if we need to to get out of a situation like that (i.e. you have seen them in Skirmish twice, play a conquest map deliberately).
#45
Posted 03 July 2014 - 07:57 AM
w0rm, on 03 July 2014 - 07:55 AM, said:
Example:
5x12 mans are in the queue: 4 of them are matched against each other, 1 of them will face a composition of smaller groups as release valve
12 man number 6 enters the queue, finds no other 12 man because 5x12 mans are allready in a match and thus will be matched against smaller groups.
Now let ELO kick in and the whole process gets even more complicated.
So depending on how fast the release valve opens (which I don't know) it is quite possible that more than one 12 man ends up facing smaller groups as long as the total number of 12 mans stays small.
More 12 mans = chances shrink that the release valve will open before they are matched against another 12 man.
Again (anecdotal) evidence: 6/7 drops the MM did put us up against smaller groups yesterday (20:00 GMT - 23:00 GMT)
Someone correct me if I'm wrong with that logic
No, that logic seems sound. I guess I was pointing out that, at any given instant, if there's an even number of 12-mans looking for a fight, they'll be paired up first.
Less 12-mans overall means you have less odd 12-mans left out to face smaller groups. I'm not saying it won't happen...it obviously will...but there aren't THAT many 12-mans out there to have anything NEAR as significant a portion of small groups facing them as people seem to be complaining about.
#46
Posted 03 July 2014 - 07:59 AM
Bilbo, on 03 July 2014 - 07:52 AM, said:
You're misreading the stats really if you think it's skewing heavily at higher groups. The sample size is sooo small, that the 8/5 match up is basically statistically closer to 50% than any other reasonable assumption.
If we played 1300 matches and saw 800/500 then I might agree with you but at this # it's not really a good sampling. As I said...anecdotal.
As I mention in a previous reply above...one of our large groups went 10 of 11 losses yesterday. Add them in the mix it's 9-15 overall, but I would wager that people are closer to 50/50 than their rage would indicate.
You have to play ALOT of games under similar conditions to get a true feel for where you are at in this game. Streaks happen.
#47
Posted 03 July 2014 - 08:24 AM
#48
Posted 03 July 2014 - 08:27 AM
Bryan Ekman, on 03 July 2014 - 08:24 AM, said:
Hey Bryan, does the MM try to match size of group first or Elo? Example: If you are in a 12 man and have another 12 man with a 400 point difference and a 8 man with a 50 point difference, will it put the 12 man against you or the 8 man?
Just curious as to the algorithm model.
#49
Posted 03 July 2014 - 08:50 AM
#50
Posted 03 July 2014 - 08:55 AM
That's at least how I understood it from the original Command Chair post a while ago. Was this idea scrapped or did I just understand it incorrectly ?
#51
Posted 03 July 2014 - 09:03 AM
Marmon Rzohr, on 03 July 2014 - 08:55 AM, said:
That's at least how I understood it from the original Command Chair post a while ago. Was this idea scrapped or did I just understand it incorrectly ?
That's the opposite of what the command post said.
Quote
http://mwomercs.com/forums/topic/162752-upcoming-patch-public-test-matchmaker-large-group-sizes-and-more/
There's no choice...it depends on what the MM needs. Solo players, however, always drop in Solo Queue.
Edited by Ghost Badger, 03 July 2014 - 09:06 AM.
#52
Posted 03 July 2014 - 09:05 AM
#53
Posted 03 July 2014 - 09:06 AM
Marmon Rzohr, on 03 July 2014 - 08:55 AM, said:
That's at least how I understood it from the original Command Chair post a while ago. Was this idea scrapped or did I just understand it incorrectly ?
you did not understand it correctly.
Few if any 2-4 mans would ever drop in group queue if they were given a choice to opt out. All these people complaining are using "choice" as a cover for what they really mean: "I want to exclusively drop in solo queue and dunk scrubs." They want to be the big fish in the small pond.
It's the same type of people who cried for an exclusive 8v8 full premade and then ran away back into pugs as soon as they realized that they would have to try.
In a competitive multiplayer game, most people will take the path of least resistance if given the choice.
#54
Posted 03 July 2014 - 09:15 AM
Davers, on 03 July 2014 - 09:05 AM, said:
Pay attention to this man, 2's can never get a choice, but the MM could be set to prefer 2's in the solo que. Statisticly I think they probably would anyways as only 10's and 7's need a 2 to fill out the 12.
#55
Posted 03 July 2014 - 09:20 AM
I'm all for a challenge and not claiming butthurt because im finding it harder. It's just this an seems extreme situation and perhaps 2-4 mans could be matched with other 2-4 mans. At the minimum I would suggest that only full 4 mans be used to make up groups to stand against 12 mans. They would at least be 3 organized lances in with a slightly better chance.
I also understand the counter argument of either play or solo or join a 12 man etc. Unfortunately not everyone is set up either domestically or whatever for teamspeak et. A casual type affair of just 1-4 friends works very well for some folk. I'm just appalled that some of my friends today - who I would consider good, competent players - are deciding that it's simply too frustating to continue playing due to the '12 man roll' effect. These are founders, and goodness knows we have all been through so many ups and downs etc.
I would ask for a bit more empathy from the 12 man side of the fence here. Try to see things from this perspective please rather than go down the hurr hurr road. After all I have fought pretty hard to fight many causes on here in my time as my post history shows.
Edited by jozkhan, 03 July 2014 - 09:21 AM.
#56
Posted 03 July 2014 - 09:22 AM
#57
Posted 03 July 2014 - 09:22 AM
Using that theory, based on what Bryan said percentage wise, the number of 12 pilot groups is relatively small. Once ELO gets re-calibrated from the MM change, 2-4 pilot groups should see less 10 and 12 pilot groups and more 4+4+4 vs 4+4+4 matches as the 8-10 and 12 group sizes rise higher in the ELO pool and are more likely to face each other.
I think the issue stems from 4 pilot groups stomping around in the solo player pool for so long they ended up with a higher ELO than their skill really dictates. So to fix the problem, play more matches, get a more realistic ELO, and the problem goes away. Either that, or there is a shortage in the player pool to allow more 4+4+4 vs 4+4+4 matches.
We were doing 10 pilot drops last night and it was the most fun I've had in a long time. There were some matches we got stomped, and some really fun close matches. The big thing for me was the "carry harder" feeling went away on a loss from solo drops. No more 800+ damage rounds that my team still ends up losing with 3-4 pilots not even hitting triple digits, or any damage at all. When we lost in our 10 pilot group, we knew why, and more importantly could do something about it next time. THAT is the biggest improvement I can see.
#58
Posted 03 July 2014 - 09:28 AM
Thanks for the info. I agree the solo queue seems ok but im not sure what you are saying as it seems contradictory.
'About 10-12% of group matches will be multiple groups vs a 12 man. Most of the time you are going up against multiple group teams (2-6 groups) depending on the time of day.'
Seems to suggest 88-90% of games are 12 man vs what exactly?
From our experience 90% of the matches we join with 1-4 players are matched vs 12 man.
#59
Posted 03 July 2014 - 09:32 AM
jozkhan, on 03 July 2014 - 09:28 AM, said:
Either vs other 12 mans or small groups vs small groups.
How in the world do you even know when you're up against a 12 man?
Edited by Charons Little Helper, 03 July 2014 - 09:32 AM.
#60
Posted 03 July 2014 - 09:37 AM
jozkhan, on 03 July 2014 - 09:28 AM, said:
Thanks for the info. I agree the solo queue seems ok but im not sure what you are saying as it seems contradictory.
'About 10-12% of group matches will be multiple groups vs a 12 man. Most of the time you are going up against multiple group teams (2-6 groups) depending on the time of day.'
Seems to suggest 88-90% of games are 12 man vs what exactly?
From our experience 90% of the matches we join with 1-4 players are matched vs 12 man.
What he's saying is that 10-12% of all group matches have a 12-man vs multiple groups, I believe.
32 user(s) are reading this topic
0 members, 32 guests, 0 anonymous users