[OFFICIAL] NetEase Explaining the Truth Behind Matchmaking

[aalbass]

Since there is a lot of speculations/misinformation in this forum,
please watch NetEase explain how matchmaking works:

Video Link: https://gdcvault.com/play/1027278/Find-the-Right-Match-AI

[Keira Lunar]

you really think im going to trust some triple A big wig ? when i have experienced first hand how the matchmaking treats me ?

[aalbass]

Originally posted by Keira Lunar:

you really think im going to trust some triple A big wig ? when i have experienced first hand how the matchmaking treats me ?

NetEase is not AAA, if you watch the video you will understand the reasons why and how the matchmaking treats you. its all controlled. Its also about psychology and churn rate

[akqpars]

Originally posted by aalbass:

Since there is a lot of misinformation in this forum,
please watch NetEase explain how matchmaking works:

Video Link: https://gdcvault.com/play/1027278/Find-the-Right-Match-AI

Thanks, it was a through and detailed explanation. It perfectly aligns with the matchmaking experience i observed.

[Paranoia_Inc]

Thanks for this video. Nice to get a lil more depth to the algorithm that either wasn't shown or was hidden amongst the formulas shown in the Optmatch document that I am simply too dumb to follow.

I guess I can kinda see why some people still think it's rigged. But they're still grossly misinterpreting what is being said. There are tricks involved to offer favorable matches to players by placing them into matches where their chances are better (As close to 51% vs 49% as it can muster) in order to better retain their play. But...that just makes your lose streaks all the more glaring and on those individuals for not cooperating and such.

That's kinda really ♥♥♥♥♥♥♥ funny.

Also starting to get an idea of what they mean by hero suppression, and is an intra-team issue. My example: ♥♥♥♥♥♥♥ Groots that keep placing walls close. These are bad, are interfering with lots of other heroes on the team, and is suppressive. It's particularly bad when the poor playing Groot is on the team meant to stop the convoys, for most of the time all they do is A) Block your team's shots just as much as the enemy team's and B) Give the escorting team a whole lot of free ground.

[aalbass]

Originally posted by Paranoia_Inc:

Thanks for this video. Nice to get a lil more depth to the algorithm that either wasn't shown or was hidden amongst the formulas shown in the Optmatch document that I am simply too dumb to follow.

I guess I can kinda see why some people still think it's rigged. But they're still grossly misinterpreting what is being said. There are tricks involved to offer favorable matches to players by placing them into matches where their chances are better (As close to 51% vs 49% as it can muster) in order to better retain their play. But...that just makes your lose streaks all the more glaring and on those individuals for not cooperating and such.

That's kinda really ♥♥♥♥♥♥♥ funny.

Also starting to get an idea of what they mean by hero suppression, and is an intra-team issue. My example: ♥♥♥♥♥♥♥ Groots that keep placing walls close. These are bad, are interfering with lots of other heroes on the team, and is suppressive. It's particularly bad when the poor playing Groot is on the team meant to stop the convoys, for most of the time all they do is A) Block your team's shots just as much as the enemy team's and B) Give the escorting team a whole lot of free ground.

Also worth mentioning is that they developed formulas for churn rate. Eight hours is the key mark, meaning if you don't play for 8 hours, the algorithm will favor you for player retention.

[tdsmith5556]

I am unable to watch the video since I am out of data and a lot of people don't feel like going through all that.

Can you just post the tl;dr of what the key points were?

[Paranoia_Inc]

Originally posted by aalbass:

Also worth mentioning is that they developed formulas for churn rate. Eight hours is the key mark, meaning if you don't play for 8 hours, the algorithm will favor you for player retention.

Was partially the eight hours and dependent on the last match's outcome. So win and then leave for eight hours and there'll be no change, and the win/loss chance would be normal. Lose and then eight hours and it'll start giving the advantages.

Thinking on it I wouldn't be surprised if matchmaking goes from the under ten seconds most everyone experiences now to something closer to 30 seconds in the next season or two as the matchmaking gets refined. Something seems a bit off about the initial phases of placing people into rooms and then rearranging them into teams to be as close to balanced as possible. I think it may be working a wee bit too fast for it to be as accurate as they want it to be and will add additional metrics.

[Paranoia_Inc]

Originally posted by tdsmith5556:

I am unable to watch the video since I am out of data and a lot of people don't feel like going through all that.

Can you just post the tl;dr of what the key points were?

Can read it here: https://nos.netease.com/mg-file/mg/neteasegamecampus/art_works/20200812/202008122020238605.pdf

[76561199506868211]

GDCvault has never been a site that was widely known before. Any data on website creation date and other things?

[tdsmith5556]

Originally posted by Paranoia_Inc:

Originally posted by tdsmith5556:

I am unable to watch the video since I am out of data and a lot of people don't feel like going through all that.

Can you just post the tl;dr of what the key points were?

Can read it here: https://nos.netease.com/mg-file/mg/neteasegamecampus/art_works/20200812/202008122020238605.pdf

I read this and didn't see anything noteable that I should care about.

Can anybody just summarize what is in that presentation we should care about regarding the matchmaking in Marvel Rivals?

[Mister Father]

Originally posted by Keira Lunar:

you really think im going to trust some triple A big wig ? when i have experienced first hand how the matchmaking treats me ?

NetEase is not AAA lmao.

[jadenedaj]

I literally speak for every single gamer when I say DELETE CHURN CALCULATIONS.
It is irrelevant to gaming.
You don't see people increasing the goal in basketball 1% because they are on a loss streak.

I didn't think this was real, I thought people were coping, but they admit to this horrible practice.

I am at a loss for words.

[Kaspar]

Originally posted by Your New Daddy:

Originally posted by Keira Lunar:

you really think im going to trust some triple A big wig ? when i have experienced first hand how the matchmaking treats me ?

NetEase is not AAA lmao.

As of early February 2025, NetEase Inc. (NTES) has a market capitalization of approximately $65.63 billion.

In comparison, other prominent AAA gaming companies have the following market capitalizations:
Electronic Arts (EA): Approximately $34.22 billion.
Take-Two Interactive (TTWO): Approximately $32.48 billion.

NetEase isn't just AAA for their quality but also for their lack of professionalism. They definitely have the resources.

That is one of the reasons I believe this game might be more about behavioral dynamics than actual profit. It would be easy for them to hire a professional community manager to work with the community to resolve the issues. However, the issues are ignored, which gives me strong reason to suspect that they might be part of the design—perhaps to see how much frustration players can endure while still remaining engaged. The insights gained from this could later be applied to marketing or even politics on a larger scale.

[pirate135246]

Originally posted by aalbass:

Originally posted by Keira Lunar:

you really think im going to trust some triple A big wig ? when i have experienced first hand how the matchmaking treats me ?

NetEase is not AAA, if you watch the video you will understand the reasons why and how the matchmaking treats you. its all controlled. Its also about psychology and churn rate

"Netease is not AAA" you just lost your entire argument

[m0n]

Yea no thanks
They have bots on their own game and refuse to acknowledge it. I won't trust any of these