Is Mm Working?
#61
Posted 03 July 2014 - 02:16 PM
Went from 2-3 kill 300-500 damage games, to 100 damage games with no a whopping 4 kills to show for getting blown up 16 times (hey, you try to do damage with 8 enemies ganged up on you because your team managed to disappear in one big affing assplosion)
#62
Posted 03 July 2014 - 02:39 PM
#63
Posted 03 July 2014 - 02:55 PM
Agent 0 Fortune, on 03 July 2014 - 02:08 PM, said:
An interesting observation, the values you used are extremely close and yet the variance number you arrived it is substantially higher than any of PGI's numbers, indicating that either the Matchmaker is extremely focused, and able to match remarkably narrow bands of Elo, or (more likely) we are seeing something else.
I can pretty much 100% guarantee you that as the picture was provided by PGI that it would be the case of narrow bands of Elo, it wouldn't make sense for them to show off anything else.
Now if the picture provided is an accurate cross section of matchmaker is anyone but PGi's guess as it could be a very good case or a standard case. I merely want people to stop complaining for the wrong reasons.
#64
Posted 03 July 2014 - 03:56 PM
Deathlike, on 02 July 2014 - 02:51 PM, said:
After playing mostly large groups yesterday, today I played solo extensively for several hours.
My experience was similar. The teams I dropped with were often of noticeably low quality. I did end up on the winning side in most of the matches; this isn't a complaint, it's an observation.
Spectating some of these guys that don't press R, don't know how to aim, go to stupid places, don't even notice enemy ECM mechs that walk in the middle of their screen in the open, and keep their machine guns and ERLLs in the same weapon group was quite something. Also, light pilots that stop before shooting.
I saw several matches in which at least 6 of my team members dealt under 50 damage each. Some of those matches we won (and not by capping). In one match, 3 non-disconnects in my team ended up dead with 0 damage (including a Timber Wolf). That's something I don't remember seeing before.
I guess most of the good players in small premades now get dragged to the group queue, likely the 8-12 mans have pretty high Elos, and they're needed there for balance. And perhaps most of the really good players are usually in premades of some sort? So the solo queue gets what's left - and it's likely much easier to find matching solo players for those not-so-great small premades.
#65
Posted 03 July 2014 - 05:30 PM
#66
Posted 03 July 2014 - 06:12 PM
Karamarka, on 03 July 2014 - 05:30 PM, said:
It's weight class.
For instance, a Locust is treated the same as a Jenner/Firestarter/Raven in terms of Elo.
Technically, it's better if it was by mech, and particular variant. For instance, most people would do better in a Raven-3L than the 2X or F4...
But... whatever.
Edited by Deathlike, 03 July 2014 - 06:12 PM.
#67
Posted 05 July 2014 - 01:04 PM
stjobe, on 02 July 2014 - 11:43 AM, said:
Edit: Well, turns out that variance actually is rather remarkably good. I stand corrected.
Well... correction again. "Elo Variance" was colloquial variance, not statistical variance.
There's a 150(-) Elo in that set. Hope that player is now getting other 100s matched up....
Karl Berg, on 05 July 2014 - 11:34 AM, said:
Hey Cimarb, unfortunately those variance numbers are not true statistical variance values. Out of lazyness I was simply calculating highest Elo on team - lowest Elo on team, rather than a true normalized sum of differences from mean.
Those screenshots were also taken within the first hour of launch of the new matchmaker. We've gone through extensive tuning passes since, both to bring wait times vs percentage 3's games in-line with stakeholder expectations, and to significantly tighten up Elo matching as well.
#68
Posted 06 July 2014 - 04:44 PM
#69
Posted 06 July 2014 - 07:17 PM
H Seldon, on 02 July 2014 - 11:23 AM, said:
No, it is not working. It seems that solo matches are weighted more toward this stupid 3/3/3/3 system rather than ELO (skill). And 3/3/3/3 is definitely failing in the group servers because the majority of matches that I have played this past weekend have been well outside of the cookie cutter drop deck being forced on us.
I hope it continues to fail and we can be rid of it for good.
#70
Posted 06 July 2014 - 07:20 PM
Dymlos2003, on 02 July 2014 - 11:28 AM, said:
While its rare that I both out damage and out kill my 11 allies, outplaying the worst two lances by a wide margin is fairly common. It must always be my fault we didn't win. I guess I've gotta accept that I have to carry a lot harder....
Just how many retar...err...bad players am I supposed to make up for anyway? More than 10 feels like a bit much to me....
#71
Posted 06 July 2014 - 07:48 PM
Matchmaker has never, ever been better.
Group queue is freaking HARDCORE mode. You drop with some friends, you play to the team and not like pugs or you gonna get the bjesus kicked out of you 4 games out of 5.
Solo queue is actually gumby mode now - the problem is that so many people don't realize it.
No premades to carry or lead either side 9 times out of 10. If they're there, they're casuals, overjoyed to not have ended up as food for the group queue beast.
USE CHAT TO COMMUNICATE WITH YOUR TEAM.
PLAY AGGRESSIVELY, DON'T MILL AROUND AND WAIT FOR SOMEONE ELSE TO PUSH FOR ROLL. PUGS RARELY RESPOND WELL TO BEING RUSHED - MOST FREEZE OR RUN. IF YOU DON'T PUSH THE OTHER TEAM LIKELY WILL.
SHOOT THE GUY SHOOTING YOUR TEAMMATE, NOT THE GUY SHOOTING YOU WHEN YOU CAN. ODDS ARE GOOD HE WILL STAND THERE AND TAKE IT LIKE AN IDJIT.
KEEP MOVING. EVEN WHEN HOLDING AN AREA RUN A FIGURE EIGHT THROUGH YOUR DEFENSIVE ZONE, GENERALLY TWO OR MORE GRIDS LONG. THIS BREAKS ATTACKERS ATTENTION FROM YOU (MOST PUGS HAVE THE ATTENTION SPAN OF A METHED UP SQUIRREL) AND PUTS YOU IN A POSITION TO ATTACK PEOPLE WHO ARE NOT EXPECTING YOU BUT ARE INSTEAD MONOFOCUSED ON YOUR TEAMMATE. QUIT POPPING FORWARD AND BACK FROM THE SAME LOCATION OR SAME AREA.
WHEN PUSHING IT IS BETTER TO RUN CLEAN PAST AN ENEMY POSITION AND INTO COVER BESIDE OR BEYOND IT THAN TO TRY TO STOP AND BACK UP. WHEN PUSHING IN HAVE AN ESCAPE ROUTE ADJACENT TO OR PAST THE GUY YOU'RE ATTACKING TO DIVERT TO. IF YOU ADVANCE USUALLY (HOPEFULLY!) YOUR TEAMMATES ARE BEHIND YOU MOVING IN AS WELL. YOU STOP AND BACK UP AND YOU BLUNT YOUR OWN ASSAULT. DON'T PLAN A PEAK AND SHOOT, PLAN A DRIVE-BY IF AT ALL POSSIBLE.
4 user(s) are reading this topic
0 members, 4 guests, 0 anonymous users