Forums / Community / Matchmaking Feedback & Discussion

[Locked] Matchmaking Feedback Update – October 16

OP ZaedynFel

  1. 1
  2. ...
  3. 2
  4. 3
  5. ...
  6. 4
No new updates this week. I’m also headed to the HCS Fall Finals at Dreamhack Denver, so next week’s update will also probably be small.

I’ll leave this thread here though to keep the communication lines open.

Hope to see some of you in Denver!
This post has been edited by a moderator. Please do not create alternate accounts to bypass forum bans. Alternate accounts will be permanently banned, and offending users will be subject to both temporary and permanent bans.
*Original post. Click at your own discretion.
Spoiler:
Show
This post has been edited by a moderator. Please do not create alternate accounts to bypass forum bans. Alternate accounts will be permanently banned, and offending users will be subject to both temporary and permanent bans.
*Original post. Click at your own discretion.
Spoiler:
Show
Is there any Idea or update about when the larger fireteams in warzone may return?
Quote:
ZaedynFelwrote:
MMR is completely different in a different playlist. The MMR of the full party was nowhere near Champ. It was good, but not amazing.

Social still has very loose matchmaking, and this match was within that tolerance.

On the other hand, the skill system knew this was going to be a blowout, but still allows it because otherwise matchmaking would take forever, which we try and avoid more in social --- though it still happens for some.

When the feature that lets us balance across teams gets fixed, this won't happen.
Yeah, I didn't think about their MMR being different, but at the time I could only view their CSR ranks which were high in slayer and FFA and that skill set would easily transfer over to a slayer game in the H3 playlist. For all I knew, they could've been playing that playlist since it came out which is why I was so annoyed by the matching, but now I know their MMR wasn't as high most likely because they haven't been playing it as much.

That was the part that initially confused me because previously you have been saying that high level players will only match other high level players, but I guess if their MMR wasn't that high in that playlist then I could see why we matched them.

I'm not sure I completely agree with that because I seemed to be finding games in a reasonable amount of time in that playlist searching focused during that time, but maybe they weren't since they were in a party? I don't know, but you would obviously know more about that over me.

Good to hear. Anyway, thanks for answering my previous question.
Could you explain why I just played 6 CTFs in a row?
We all know it... i'm just putting it out there again .........

Is it me or have the smurfs totally ruined this game for Gold and platinum rank players. .... I mean comon sr10 and getting 25 kills in the gold bracket but your rank is bronze 6... and this use to happen maybe 1 out of 10 games it seems to be every other game anymore with Sr10 to Sr148 they are everywhere ... yes smurfs have always been around for ever i know that ... but halo 5 is so easy to smurf on ....... its really gotten out of hand i have cut my play time this season to only a few games a week and only ranked one play list slayer because of this..and how bad they are this season i'm probably going cut more play time until you guys can move forward and at least slow them down some ... you shouldn't get wrecked every game
im just a average play I'm not no Savage i rank around gold 4 or 5 and by end of season i use to be able to make plat 5 or 6 playing SOLO haven't been able do that for the last few seasons... this season i ranked gold 6 and with people not playing (smurfing a new or old account) or the other team has one or two smurfs on i'm down to gold 3 and cant seem to rank up anymore ... I shouldn't have to rally up a team ever time i want to play this game .......... just what to play people at skill level you guys have put me at
jzm0066 wrote:
Could you explain why I just played 6 CTFs in a row?
Because your luck is high.
jzm0066 wrote:
Could you explain why I just played 6 CTFs in a row?
Are you backing out between matches or waiting through intermission?
Is there any Idea or update about when the larger fireteams in warzone may return?
Not yet, but the last I heard the team working on that has made good progress. It's just not part of 343 so I try not to speak too much for them.
jzm0066 wrote:
Could you explain why I just played 6 CTFs in a row?
Random luck. The modes are equally likely, so it can happen.
Quote:
And as to your question about comms, yes, we still have millions of data points showing players on comms losing to randos. That skill gap is just as quantifiable.
@zaedynfel

do your data points differentiate between comm callouts like “ugh what!” Vs “one shot my x” vs “2 shots p2?” Because those are very different callouts from people with very different skill levels.

also do your data points show the timing of those callouts? There is a huge difference between calling out “2 shots pink 2” during a fight vs 1 second after you’re dead.

Quote:
You shouldn't get to win more because you're in a team.
Uhhhh, what? What is this world? We’re obviously 2 fundamentally different people who will never agree on this. That said, I have a couple of new questions.

1. How do smurfs effect these team vs rando stats?
2. How do Smurfs effect your 50/50 goal and why aren’t they the focus of your efforts? I’m running into game ruining smurfs more than 25% of the time now.
jzm0066 wrote:
Could you explain why I just played 6 CTFs in a row?
Are you backing out between matches or waiting through intermission?
ZaedynFel wrote:
Random luck. The modes are equally likely, so it can happen.
Backing out mostly. What about maps 3 out 6 were fathom with 2 in a row? I was playing a lot of the same people too so shouldn't there be more variety.
I ran into, what I think, is a new kind of smurf yesterday.

A silver 3, with a 100% win percentage, and a kda of 11.5. I’ve seen Smurf’s with a high kda but never one with a 100% win rate and a silver 3 rank. How is this possible? How do you win all of your qualifiers and finish with a silver 3 rank?
jzm0066 wrote:
jzm0066 wrote:
Could you explain why I just played 6 CTFs in a row?
Are you backing out between matches or waiting through intermission?
ZaedynFel wrote:
Random luck. The modes are equally likely, so it can happen.
Backing out mostly. What about maps 3 out 6 were fathom with 2 in a row? I was playing a lot of the same people too so shouldn't there be more variety.
Backing out resets your hopper. If you stay through intermission the system remembers what you've played and tries to vary things more.
SGO SMACK wrote:
I ran into, what I think, is a new kind of smurf yesterday.

A silver 3, with a 100% win percentage, and a kda of 11.5. I’ve seen Smurf’s with a high kda but never one with a 100% win rate and a silver 3 rank. How is this possible? How do you win all of your qualifiers and finish with a silver 3 rank?
A thought comes to mind:

  • The Silver was playing with a group/party of Bronze-Silver players which meant that their team MMR average would ensure that he/she was more likely to face off against opponents with a similar team MMR level (in the Bronze-Silver range). Due to the fact that the player may not be facing opponents of greater skill then what they themselves possess their MMR/CSR boost after winning matches is quite limited. Perhaps the player may be a high Silver or low Gold level player but their Rank is somewhat affected by the restrictions brought forth by those who they are playing with and against. Supposedly, the first few matches a player ever plays within a playlist will have the matchmaker taking a look at their total performance (K/D) in order to more quickly adjust their MMR levels; however, if they perform poorly within those first few matches, but hit their groove afterwards then their MMR/CSR will be mostly impacted by those they're playing with and against -- maybe that's the case with the person you're referencing.
From my understanding the redesigned TrueSkill system that's in the pipeline for Halo should help clean some of this up and more accurately indicate their individual skill level.
SGO SMACK wrote:
Quote:
And as to your question about comms, yes, we still have millions of data points showing players on comms losing to randos. That skill gap is just as quantifiable.
@zaedynfel

do your data points differentiate between comm callouts like “ugh what!” Vs “one shot my x” vs “2 shots p2?” Because those are very different callouts from people with very different skill levels.

also do your data points show the timing of those callouts? There is a huge difference between calling out “2 shots pink 2” during a fight vs 1 second after you’re dead.

Quote:
You shouldn't get to win more because you're in a team.
Uhhhh, what? What is this world? We’re obviously 2 fundamentally different people who will never agree on this. That said, I have a couple of new questions.

1. How do smurfs effect these team vs rando stats?
2. How do Smurfs effect your 50/50 goal and why aren’t they the focus of your efforts? I’m running into game ruining smurfs more than 25% of the time now.
We don't do speech recognition on call outs, but we can observe how big of a difference being in a party makes relative to your skill level. We can tell whether, for example, being in a party results in a bigger advantage in Onyx vs. Onyx than it does in Gold vs. Gold. Current results suggest the difference isn't significant enough to account for, though we are still evaluating that.

This doesn't mean higher skilled teams aren't more coordinated overall, it just suggests that, relative to their skill levels, the advantage is the same. It suggests that Onyx solo players are so good that the advantage stays about the same.

It's still a significant advantage either way (if all of the players are exactly the same skill, the party will win 3 of 4).

For those who like the math, the advantage is close to 0.5 MMR per player, or 2.0 MMR total in 4v4, for CTF and Strongholds (it's less for Slayer since that mode requires less teamwork). 2.0 MMR is a fairly large gap between teams (75/25).

As for smurfs, they happen a lot less than their perception influences, so they overall don't have a huge impact on the overall system.

But, they are a super negative experience, which makes them feel like they are happening orders of magnitude more than they really are.

This same new system actually handles smurfs very well, and they stand out pretty clearly, so we'll be able to reduce their impact on player experiences significantly.
ZaedynFel wrote:
SGO SMACK wrote:
Quote:
And as to your question about comms, yes, we still have millions of data points showing players on comms losing to randos. That skill gap is just as quantifiable.
@zaedynfel

do your data points differentiate between comm callouts like “ugh what!” Vs “one shot my x” vs “2 shots p2?” Because those are very different callouts from people with very different skill levels.

also do your data points show the timing of those callouts? There is a huge difference between calling out “2 shots pink 2” during a fight vs 1 second after you’re dead.

Quote:
You shouldn't get to win more because you're in a team.
Uhhhh, what? What is this world? We’re obviously 2 fundamentally different people who will never agree on this. That said, I have a couple of new questions.

1. How do smurfs effect these team vs rando stats?
2. How do Smurfs effect your 50/50 goal and why aren’t they the focus of your efforts? I’m running into game ruining smurfs more than 25% of the time now.
We don't do speech recognition on call outs, but we can observe how big of a difference being in a party makes relative to your skill level. We can tell whether, for example, being in a party results in a bigger advantage in Onyx vs. Onyx than it does in Gold vs. Gold. Current results suggest the difference isn't significant enough to account for, though we are still evaluating that.

This doesn't mean higher skilled teams aren't more coordinated overall, it just suggests that, relative to their skill levels, the advantage is the same. It suggests that Onyx solo players are so good that the advantage stays about the same.

It's still a significant advantage either way (if all of the players are exactly the same skill, the party will win 3 of 4).

For those who like the math, the advantage is close to 0.5 MMR per player, or 2.0 MMR total in 4v4, for CTF and Strongholds (it's less for Slayer since that mode requires less teamwork). 2.0 MMR is a fairly large gap between teams (75/25).

As for smurfs, they happen a lot less than their perception influences, so they overall don't have a huge impact on the overall system.

But, they are a super negative experience, which makes them feel like they are happening orders of magnitude more than they really are.

This same new system actually handles smurfs very well, and they stand out pretty clearly, so we'll be able to reduce their impact on player experiences significantly.
Being in a party or having comms doesn’t tell any real story. All too often people have comms but don’t use them. Those who do, use them VERY differently. The vast majority of callouts I hear are almost useless.

Am I running into more smurfs than the average player because of my rank? Because it happens a lot. It’s not just my perception. It’s constant in hcs and doubles. Doubles is the worst.

i truly hope this new system renders smurfs useless in ranked play. They destroy rankings. It’s bad enough I have to lose to them, it’s even worse that they sometimes drop me 3/4 of a level because of their low rank.
eLantern wrote:
SGO SMACK wrote:
I ran into, what I think, is a new kind of smurf yesterday.

A silver 3, with a 100% win percentage, and a kda of 11.5. I’ve seen Smurf’s with a high kda but never one with a 100% win rate and a silver 3 rank. How is this possible? How do you win all of your qualifiers and finish with a silver 3 rank?
A thought comes to mind:

  • The Silver was playing with a group/party of Bronze-Silver players which meant that their team MMR average would ensure that he/she was more likely to face off against opponents with a similar team MMR level (in the Bronze-Silver range). Due to the fact that the player may not be facing opponents of greater skill then what they themselves possess their MMR/CSR boost after winning matches is quite limited. Perhaps the player may be a high Silver or low Gold level player but their Rank is somewhat affected by the restrictions brought forth by those who they are playing with and against. Supposedly, the first few matches a player ever plays within a playlist will have the matchmaker taking a look at their total performance (K/D) in order to more quickly adjust their MMR levels; however, if they perform poorly within those first few matches, but hit their groove afterwards then their MMR/CSR will be mostly impacted by those they're playing with and against -- maybe that's the case with the person you're referencing.
From my understanding the redesigned TrueSkill system that's in the pipeline for Halo should help clean some of this up and more accurately indicate their individual skill level.
He was in doubles with a diamond teamate. He ranked his 10th game against me. I appreciate the theory but, I don’t think that’s what happened.
SGO SMACK wrote:
eLantern wrote:
SGO SMACK wrote:
I ran into, what I think, is a new kind of smurf yesterday.

A silver 3, with a 100% win percentage, and a kda of 11.5. I’ve seen Smurf’s with a high kda but never one with a 100% win rate and a silver 3 rank. How is this possible? How do you win all of your qualifiers and finish with a silver 3 rank?
A thought comes to mind:

  • The Silver was playing with a group/party of Bronze-Silver players which meant that their team MMR average would ensure that he/she was more likely to face off against opponents with a similar team MMR level (in the Bronze-Silver range). Due to the fact that the player may not be facing opponents of greater skill then what they themselves possess their MMR/CSR boost after winning matches is quite limited. Perhaps the player may be a high Silver or low Gold level player but their Rank is somewhat affected by the restrictions brought forth by those who they are playing with and against. Supposedly, the first few matches a player ever plays within a playlist will have the matchmaker taking a look at their total performance (K/D) in order to more quickly adjust their MMR levels; however, if they perform poorly within those first few matches, but hit their groove afterwards then their MMR/CSR will be mostly impacted by those they're playing with and against -- maybe that's the case with the person you're referencing.
From my understanding the redesigned TrueSkill system that's in the pipeline for Halo should help clean some of this up and more accurately indicate their individual skill level.
He was in doubles with a diamond teamate. He ranked his 10th game against me. I appreciate the theory but, I don’t think that’s what happened.
I'd have to see the Silver's profile.

There are a few ways this can happen, but I'm only guessing without a link to them.

For example, if the player played poorly in past seasons and has a Silver MMR, and somehow between seasons became an amazing player, he would be matching Silvers his first few matches and destroying them. There are legitimate and illegitimate ways this can happen.
  1. 1
  2. ...
  3. 2
  4. 3
  5. ...
  6. 4