Marvel Rivals’ Matchmaking: The Deepest Investigation You’ll Ever See

00:55:37
https://www.youtube.com/watch?v=tRHnsd9A8Cc

概要

TLDRThis investigative video delves into the matchmaking system of Marvel Rivals, examining perceptions of whether it is rigged or not. The speaker traces the history of online game matchmaking, from simple lobby systems to advanced algorithms like skill-based matchmaking (SBMM) and engagement optimized matchmaking (EOMM). Player experiences are analyzed, along with research papers detailing matchmaking processes, revealing how systems might nudge win rates to promote engagement rather than pure competition. The conclusion indicates that while matchmaking incorporates strategic elements to enhance player satisfaction, it may sometimes feel against the principles of fair competition.

収穫

  • 🕹️ Matchmaking has evolved significantly from simple systems to advanced algorithms.
  • 🏆 EOMM prioritizes engagement, potentially over strict fairness.
  • 📈 Many players believe their win rates are manipulated to average around 50%.
  • 🔍 Research and data suggest the presence of complex factors in matchmaking nuances.
  • 🎮 Transparency in matchmaking algorithms can alleviate player suspicions.

タイムライン

  • 00:00:00 - 00:05:00

    In this introduction, the speaker discusses their journey in Marvel Rivals, experiencing both winning and losing streaks. They ponder whether these experiences are due to bad luck or a manipulated matchmaking system and mention their extensive research into online gaming matchmaking algorithms.

  • 00:05:00 - 00:10:00

    The speaker reflects on the history of matchmaking in online games, explaining the early days where fair matches weren't prioritized. They introduce the concept of skill rating systems inspired by chess and the evolution of matchmaking algorithms, highlighting the improvements made to create more balanced matches.

  • 00:10:00 - 00:15:00

    The discussion continues with the evolution of skill-based matchmaking systems (SBMM) and introduces TrueSkill as a notable development that aimed to create fair fights in games. Players were matched based on their demonstrated skill, avoiding mismatched skill levels.

  • 00:15:00 - 00:20:00

    The speaker notes that, as the gaming landscape evolved, developer focus shifted toward player retention. They introduce Engagement Optimized Matchmaking (EOMM), emphasizing that the aim was not just fair competition but to keep players engaged in the game.

  • 00:20:00 - 00:25:00

    They explain how EOMM leverages player data such as churn probability and dynamic adjustments to maximize engagement, likening the matchmaking system to a casino dealer who may tweak outcomes subtly to maintain player interest instead of strictly adhering to skill-based principles.

  • 00:25:00 - 00:30:00

    The speaker highlights how major game companies, including Activision, have also sought to optimize matchmaking not only for fair play but to encourage player spending through strategic pairings. Marvel Rivals, being a newer game, is introduced as having a research-focused matchmaking design influenced by EOMM principles.

  • 00:30:00 - 00:35:00

    The speaker posits that Marvel Rivals may not follow pure SBMM, but an adaptation called 'opt match' that seeks to maximize player satisfaction, balancing skill parity with engagement metrics. Details about further research into matchmaking dynamics are set to follow.

  • 00:35:00 - 00:40:00

    They delve into the specifics of the 'opt match' system, explaining its two-phase approach: offline learning for analyzing player performance and online matchmaking planning that considers player and hero dynamics to create satisfying matchups.

  • 00:40:00 - 00:45:00

    As the talk progresses, they describe how the matchmaking system attempts to create balanced teams by analyzing player history and gameplay styles, aiming to enhance both competitiveness and enjoyment of matches without predetermined outcomes.

  • 00:45:00 - 00:50:00

    The speaker shares community feedback on matchmaking perceptions, detailing claims of a 'rigged' system as players note an unnaturally consistent 50% win rate, feelings of being paired with weaker teammates, and theories of hidden mechanics to influence game outcomes.

  • 00:50:00 - 00:55:37

    Finally, the video examines the criticisms of Marvel Rivals' matchmaking including perceived unfairness and frustration from high-skilled players who feel that engagement-driven designs hinder pure competitive experiences, posing questions about player transparency and expectations for future online gaming.

もっと見る

マインドマップ

ビデオQ&A

  • What is engagement optimized matchmaking (EOMM)?

    EOMM is a matchmaking approach that prioritizes player engagement over strictly fair matches, aiming to keep players interested and playing longer.

  • Why do players feel the matchmaking is rigged?

    Players often report that their win rates gravitate around 50%, experiencing abrupt winning and losing streaks, which contributes to the feeling of manipulation.

  • How has matchmaking evolved in online gaming?

    Matchmaking has evolved from simple lobby systems to complex algorithms like SBMM and EOMM, aiming for fairer and more engaging matches.

  • What factors does Marvel Rivals' matchmaking consider?

    It considers player skill, recent performance, team composition, and engagement metrics to optimize match balance.

  • Does overpowering a player or using bots affect matchmaking outcomes?

    There is no concrete evidence that artificial adjustments like nerfing players' damage or including bots are used to influence match outcomes.

ビデオをもっと見る

AIを活用したYouTubeの無料動画要約に即アクセス！
字幕
en
オートスクロール:
  • 00:00:00
    imagine logging into Marvel Rivals you
  • 00:00:02
    got your chippies you got your Cola you
  • 00:00:05
    got everything going and you know what
  • 00:00:08
    you're crushing match after match
  • 00:00:10
    Victory after Victory the streak is real
  • 00:00:14
    but just when you're riding High the
  • 00:00:17
    tide turns suddenly you find yourself
  • 00:00:20
    teamed with a bunch of noobs and your
  • 00:00:22
    opponents they seem like freaking
  • 00:00:25
    experts was it bad luck or was it
  • 00:00:27
    something planned this roller coaster of
  • 00:00:31
    emotions is not organic some players
  • 00:00:34
    insist it's engineered in this
  • 00:00:36
    investigative Deep dive that I have
  • 00:00:39
    created this monstrosity of research
  • 00:00:42
    we're going to unravel the mystery I'm
  • 00:00:44
    going to take you on a journey buddy
  • 00:00:46
    we're first going to Journey Through the
  • 00:00:48
    history of online game matchmaking and
  • 00:00:51
    then we're going to discover how modern
  • 00:00:53
    algorithms like eomm and opt match don't
  • 00:00:56
    worry those things are going to make
  • 00:00:57
    sense work behind the scenes and why
  • 00:01:00
    many players feel the system is stacking
  • 00:01:03
    the deck and hey for the last two weeks
  • 00:01:05
    I've been going blind on research papers
  • 00:01:08
    and technical things and some bits of
  • 00:01:11
    code at some point and I've really
  • 00:01:13
    intended to make a very complete well
  • 00:01:16
    researched wellth thought out video here
  • 00:01:19
    so your likes and comments and
  • 00:01:20
    subscriptions are deeply appreciated as
  • 00:01:22
    you can see I have less than 1,000
  • 00:01:24
    subscribers and yet I'm putting in all
  • 00:01:26
    this effort and it makes me feel like an
  • 00:01:28
    idiot a little bit but at the same time
  • 00:01:31
    couldn't catch me dead just putting on
  • 00:01:32
    gameplay for 40 minutes and talking like
  • 00:01:34
    an idiot so if you want to reward me for
  • 00:01:37
    my effort here you know how to do it all
  • 00:01:39
    right let's jump into the video history
  • 00:01:42
    time see I thought that to understand
  • 00:01:45
    Marvel Rivals as matchmaking I really
  • 00:01:47
    needed to first see how matchmaking and
  • 00:01:50
    online gaming evolved and and how it all
  • 00:01:53
    started and you know what it wasn't
  • 00:01:55
    always Guided by these complex crazy
  • 00:01:58
    algorithms and that's sounds right when
  • 00:02:00
    I was playing Counter Strike as a kid I
  • 00:02:02
    don't think there was much behind the
  • 00:02:03
    scenes so in the early days of online
  • 00:02:06
    gaming matchmaking really just meant
  • 00:02:10
    just a simple Lobby or a server where
  • 00:02:13
    anybody could join and that's how I
  • 00:02:15
    remember it that's what I grew up on
  • 00:02:17
    that's probably what you grew up on
  • 00:02:18
    we're both old so you remember it it was
  • 00:02:21
    the Wild West man sometimes you're going
  • 00:02:23
    to be facing Pros way above your level
  • 00:02:26
    and just accept the loss and other times
  • 00:02:28
    you're going to be running people over
  • 00:02:30
    because it's just a noob that joined
  • 00:02:32
    there was a push for fair matches right
  • 00:02:34
    so this push led to skill rating systems
  • 00:02:38
    a really good example that I found is
  • 00:02:40
    the ELO rating system from chess of all
  • 00:02:42
    things which inspired early game
  • 00:02:44
    rankings I thought that was pretty cool
  • 00:02:46
    so by 2004 the developers of Halo 2 were
  • 00:02:50
    messing around with the new Bayesian
  • 00:02:52
    rating called true skill to improve upon
  • 00:02:55
    ELO and I think that's where it all
  • 00:02:57
    starts see true skill aimed to Qui
  • 00:02:59
    quickly estimate a player's skill and
  • 00:03:02
    then match them up with others of
  • 00:03:04
    similar ability making the matches a bit
  • 00:03:06
    more balanced throughout uh 2000s and
  • 00:03:08
    2000s and 10s a lot of competitive games
  • 00:03:11
    like League of
  • 00:03:13
    Legends ad Dota 2 adopted such
  • 00:03:17
    skill-based matchmaking systems by the
  • 00:03:19
    way this skill-based matchmaking thing
  • 00:03:21
    it's just we're going to call it sbmm
  • 00:03:24
    because that abbreviation is going to
  • 00:03:25
    show up a lot later in the video so just
  • 00:03:27
    remember it skill-based matchmaking say
  • 00:03:30
    it with me and now this spmm it led to
  • 00:03:33
    this matchmaking rating or it was often
  • 00:03:36
    tied to it which is MMR which is
  • 00:03:37
    probably something you're familiar with
  • 00:03:39
    and that rises with your wins and losses
  • 00:03:42
    your MMR right got me skill best
  • 00:03:44
    matchmaking it takes MMR they do this or
  • 00:03:49
    maybe this and then you got true skill
  • 00:03:51
    that's basically what it was Story Time
  • 00:03:54
    picture a 2007 Halo 3 Lobby back in the
  • 00:03:59
    peak of delicious racism the game's spmm
  • 00:04:04
    quietly crunches numbers like a coach
  • 00:04:06
    figuring out which players to let play
  • 00:04:09
    so if you're a sharpshooter in the last
  • 00:04:11
    games and you had perfect aim and you
  • 00:04:12
    just killed everybody the system is then
  • 00:04:14
    going to bump up your invisible skill
  • 00:04:16
    number a bit and then next match it's
  • 00:04:19
    going to try to seat you at the table
  • 00:04:20
    with players of comparable aim and the
  • 00:04:23
    goal of all of that was really simple
  • 00:04:25
    just fair fights so at its score sbm M
  • 00:04:30
    treated matchmaking like a chess
  • 00:04:32
    tournament so you're going to match
  • 00:04:34
    Grand masters with grand Masters and
  • 00:04:36
    you're going to match novices with
  • 00:04:37
    novices pretty simple but as online
  • 00:04:40
    gaming grew and expanded a new question
  • 00:04:43
    came up what if Fair fights are not the
  • 00:04:46
    only goal so naturally game companies
  • 00:04:49
    started to notice something player
  • 00:04:52
    retention you know how long you keep
  • 00:04:54
    playing is a fragile thing and it kind
  • 00:04:58
    of makes sense you know a perfectly Fair
  • 00:05:00
    system might inadvertently cause some
  • 00:05:02
    players to quit so then of course
  • 00:05:05
    developers started to Wonder Could
  • 00:05:07
    matchmaking be tuned not just for
  • 00:05:09
    fairness but to keep players engaged see
  • 00:05:12
    this led to the rise of Engagement
  • 00:05:14
    optimized matchmaking which was the
  • 00:05:16
    Natural Evolution from sbmm to this or
  • 00:05:20
    eomm engagement optimized matchmaking
  • 00:05:22
    see in 2017 a team at Electronic Arts
  • 00:05:25
    published a paper titled eomm engagement
  • 00:05:29
    optimized match making framework it was
  • 00:05:32
    a really a watershed moment the paper
  • 00:05:34
    suggested that perfectly Fair matches
  • 00:05:36
    are not always the best for keeping
  • 00:05:38
    players hooked instead of only using
  • 00:05:41
    skill the eomm framework added new
  • 00:05:44
    ingredients to the matchmaking recipe
  • 00:05:46
    the first one was churn probability this
  • 00:05:48
    is an estimate of How likely you are to
  • 00:05:50
    quit playing if you have a bad or good
  • 00:05:53
    experience the second one were Dynamic
  • 00:05:55
    adjustments so the system could even
  • 00:05:58
    change its objective which which is nuts
  • 00:06:00
    the researchers then noted that they
  • 00:06:02
    could retune the matchmaking goal to
  • 00:06:05
    maximize other things like playtime
  • 00:06:08
    retention or even spending which is when
  • 00:06:10
    it gets real creepy to dumb this down a
  • 00:06:13
    little bit so it's a little bit easier
  • 00:06:15
    to digest eomm was basically matchmaking
  • 00:06:18
    meeting data science for the first time
  • 00:06:21
    ever so here's here's how I like
  • 00:06:22
    thinking about it I imagine the the
  • 00:06:24
    matchmaking system as as a casino dealer
  • 00:06:27
    who doesn't just deal cards randomly but
  • 00:06:29
    also keeps an eye on every player's mood
  • 00:06:31
    so if for example you're getting upset
  • 00:06:33
    and might leave the table the dealer
  • 00:06:36
    then might just tweak the next round in
  • 00:06:38
    your favor a little bit not enough to be
  • 00:06:40
    obvious but enough to keep you in the
  • 00:06:42
    game or if you're winning too much and
  • 00:06:44
    you're starting to get bored maybe the
  • 00:06:46
    next round he gives you a bit more
  • 00:06:48
    challenge so it's not I don't think it's
  • 00:06:49
    outright cheating the cards are still
  • 00:06:52
    you know dealt from the same deck but
  • 00:06:54
    the order of the cards might be
  • 00:06:55
    influenced by another motive keep the
  • 00:06:58
    game going that's kind of what's going
  • 00:07:00
    on here so around the same time other
  • 00:07:02
    gaming Giants are starting to wake up
  • 00:07:04
    and are starting to explore very similar
  • 00:07:06
    ideas in fact Activision they made Call
  • 00:07:09
    of Duty filed a patent for a matchmaking
  • 00:07:11
    system that could encourage
  • 00:07:13
    microtransactions by pairing players
  • 00:07:16
    strategically
  • 00:07:18
    wow sounds like Activision to be honest
  • 00:07:21
    so here here's how it worked for example
  • 00:07:23
    they'll be matching a newbie with a
  • 00:07:26
    veteran who has a bunch of Premium cool
  • 00:07:28
    gear and then hope that the newb is
  • 00:07:31
    going to be enticed to buy that gear
  • 00:07:33
    what the [ __ ] the industry clearly was
  • 00:07:37
    interested in matchmaking as more than
  • 00:07:39
    just Fair competition it was becoming a
  • 00:07:42
    tool to drive player Behavior enter
  • 00:07:44
    Marvel Rivals see the game is still very
  • 00:07:47
    very young as of the making of this
  • 00:07:49
    video we're in season 1 so it's still a
  • 00:07:51
    baby and the game came onto the scene a
  • 00:07:54
    few months ago with really fun hero
  • 00:07:56
    battles a great IP fantastic animation
  • 00:07:59
    and sound design and This Promise of a
  • 00:08:02
    really fun competitive team play but
  • 00:08:05
    behind the Marvel heroes is a developer
  • 00:08:07
    with a very research oriented approach
  • 00:08:10
    to matchmaking and that's net e their
  • 00:08:12
    involvement is really crucial because
  • 00:08:15
    they have been at the Forefront of
  • 00:08:16
    implementing engagement based
  • 00:08:18
    matchmaking systems in their games
  • 00:08:20
    Guided by research like opt match and
  • 00:08:22
    end match which we're going to talk
  • 00:08:23
    about in a second now let's break down
  • 00:08:26
    what I already know and suspect about
  • 00:08:28
    Marvel rivals is matchmaking the very
  • 00:08:30
    first thing is that Marvel Rivals does
  • 00:08:32
    not appear to have a traditional you
  • 00:08:35
    know pure sbmm where only skill matters
  • 00:08:40
    instead a lot of evidence suggests that
  • 00:08:41
    it uses a system influenced by eomm
  • 00:08:45
    principles potentially it's an in-house
  • 00:08:47
    variant that they call opt match then
  • 00:08:49
    the goal of such a system is just to
  • 00:08:52
    balance matches in a way that optimizes
  • 00:08:55
    and maximizes player satisfaction and
  • 00:08:57
    engagement not just skill skill parody
  • 00:09:00
    which is very different next we're about
  • 00:09:02
    to get nerdy and um I I really like
  • 00:09:04
    learning about these things but we're
  • 00:09:06
    going to talk about kind of the science
  • 00:09:08
    behind the scenes eomm opt match and
  • 00:09:11
    things like end match they'll become
  • 00:09:13
    clear as as I talk so to to roll the
  • 00:09:16
    tape back a bit in August 2020 net e
  • 00:09:19
    researchers published op matched
  • 00:09:21
    optimizing matchmaking via modeling the
  • 00:09:23
    high order interactions on the arena
  • 00:09:26
    what this was was essentially um an
  • 00:09:28
    engagement oriented matchmaking system
  • 00:09:31
    but with a really interesting twist it
  • 00:09:33
    still valued Fair close games as a route
  • 00:09:35
    to happy players but opt match also
  • 00:09:38
    introduced a two-phase approach number
  • 00:09:40
    one was offline learning so what it does
  • 00:09:42
    is that it analyzes a ton of game data
  • 00:09:45
    to understand things like composition
  • 00:09:48
    and player performance things like that
  • 00:09:50
    the paper then describes features like
  • 00:09:52
    hero to VEC and player to VC basically
  • 00:09:55
    learning the characteristics of Heroes
  • 00:09:57
    and players so things like who works
  • 00:10:00
    well with whom uh which Heroes synergize
  • 00:10:03
    so Iron Man amplifying Hulk or one
  • 00:10:07
    character covering another one's
  • 00:10:08
    weakness you know also it thinks about
  • 00:10:10
    how does this player usually perform and
  • 00:10:13
    what heroes are they best with so that's
  • 00:10:15
    the first one the second one is online
  • 00:10:18
    matchmaking planning so with all of this
  • 00:10:20
    knowledge when you hit find match the
  • 00:10:22
    system isn't just you know randomly
  • 00:10:24
    throwing six players into two teams and
  • 00:10:26
    calling it a day no it's thinking it's
  • 00:10:28
    evaluating like thousands of possible
  • 00:10:31
    team combinations among players of
  • 00:10:33
    pretty much the same rank so it's as if
  • 00:10:36
    the matchmaking is playing like a giant
  • 00:10:39
    game of Tetris fitting different pieces
  • 00:10:41
    together which are actually the players
  • 00:10:43
    into two balanced teams yeah it uses a
  • 00:10:45
    utility function which is essentially a
  • 00:10:47
    score for how satisfying a matchup is
  • 00:10:50
    predicted to be for each player and then
  • 00:10:52
    tries to maximize that if putting you
  • 00:10:55
    and player B on the same team let's say
  • 00:10:58
    because your favorite heres compliment
  • 00:11:00
    each other or something leads to a
  • 00:11:01
    higher satisfaction score the system
  • 00:11:04
    leans that way however if grouping I'll
  • 00:11:07
    just give you a bad example but grouping
  • 00:11:09
    too many tanks or too many supports of
  • 00:11:11
    the same role makes the team worse then
  • 00:11:13
    it's going to avoid that so op matches
  • 00:11:15
    philosophy in 2020 was still Fair
  • 00:11:18
    matches equals happy players at this
  • 00:11:21
    stage it wasn't you know rigging
  • 00:11:23
    outcomes it was trying to make matches
  • 00:11:25
    as fair and as fun as possible by really
  • 00:11:28
    smartly considering hero Dynamics and
  • 00:11:31
    player play Styles which I think is
  • 00:11:32
    pretty cool I'll give you a metaphor
  • 00:11:34
    here to help you digest this a bit more
  • 00:11:36
    it's like a sports coach who not only
  • 00:11:38
    matches teams by skill but also thinks
  • 00:11:41
    okay this team has all Strikers and no
  • 00:11:43
    goalkeepers so that's going to be
  • 00:11:45
    pointless and not fun so let's swap it
  • 00:11:48
    around a bit to balance roles however
  • 00:11:50
    later developments show that net ease
  • 00:11:52
    didn't just stopped there by December
  • 00:11:54
    2020 the same researchers put out a
  • 00:11:57
    GitHub simulation testing different m
  • 00:11:59
    matchmaking algorithms including eomm
  • 00:12:02
    it's way too nerdy it didn't make any
  • 00:12:03
    sense but it was pretty cool to see so
  • 00:12:05
    at a game developers conference in 2021
  • 00:12:07
    or GDC net e detailed how they started
  • 00:12:11
    integrating this churn prediction and
  • 00:12:13
    player streaks into the equation and
  • 00:12:15
    this is where things get intriguing and
  • 00:12:18
    kind of controversial I think so they
  • 00:12:21
    introduced this concept of player
  • 00:12:23
    utility that included not just fairness
  • 00:12:26
    but also match duration cu time and and
  • 00:12:30
    recent win loss streaks for example if
  • 00:12:32
    you've been on a losing streak the
  • 00:12:34
    system is going to take that into
  • 00:12:35
    account your utility for example how
  • 00:12:38
    happy you'll be in the next match might
  • 00:12:40
    be lower because you're frustrated so
  • 00:12:41
    it's just going to compute all of that
  • 00:12:43
    then the matchmaking might respond by
  • 00:12:46
    trying to give you a better experience
  • 00:12:48
    the next round to keep your spirits up
  • 00:12:51
    more specifically that could mean aiming
  • 00:12:53
    for a match that's predicted to be close
  • 00:12:55
    and exciting or ensuring to get you into
  • 00:12:58
    the game quicker or you know maybe
  • 00:13:00
    balancing the teams in such a way that
  • 00:13:02
    you have a fair shot despite recent
  • 00:13:05
    losses in that GDC talk one slide
  • 00:13:07
    specifically contained a really
  • 00:13:09
    interesting line give tiny advantages to
  • 00:13:12
    the desperate or fatigued players during
  • 00:13:15
    matchmaking if possible for example
  • 00:13:18
    choose the win rate of 51% instead of
  • 00:13:20
    49% matchmaking proposals for those
  • 00:13:23
    players this basically implies that if
  • 00:13:26
    the system has a choice between two
  • 00:13:29
    possible matchups one where it predicts
  • 00:13:31
    you have a 49% chance to win and one
  • 00:13:35
    where you have predicts you have a 51%
  • 00:13:37
    chance to win and you've been on a rough
  • 00:13:39
    losing streak so by by their definition
  • 00:13:41
    you're maybe desperate or fatigued it
  • 00:13:43
    will gently nudge you towards the 51%
  • 00:13:47
    scenario in other words it's becoming a
  • 00:13:50
    slight kind of hidden handicap in your
  • 00:13:52
    favor just to help you out on the flip
  • 00:13:54
    side it also made me think what about
  • 00:13:56
    winning streaks so the research doesn't
  • 00:13:59
    really like explicitly say give
  • 00:14:02
    disadvantages to players on winning
  • 00:14:04
    streaks however if the system is always
  • 00:14:07
    nudging losing streak players upward a
  • 00:14:10
    bit then you know the logical
  • 00:14:12
    consequence is that if you are on a
  • 00:14:14
    winning streak you might lose that
  • 00:14:17
    gentle nudge makes sense right so you're
  • 00:14:19
    essentially left to face the full force
  • 00:14:21
    of competition or maybe even you know
  • 00:14:24
    given a slightly tougher matchup if
  • 00:14:26
    others are being boosted and and this
  • 00:14:28
    could naturally result in your streak
  • 00:14:29
    coming to an end sooner or later not
  • 00:14:32
    because the game hates you but because
  • 00:14:34
    it just stops giving you training wheels
  • 00:14:37
    so net EAS continued to refine this
  • 00:14:40
    approach time and time again in a 2024
  • 00:14:42
    paper they very frankly state that the
  • 00:14:45
    traditional Fair matchmaking you know
  • 00:14:47
    the sbmm thing dividing players strictly
  • 00:14:50
    by skill tiers is not always good for
  • 00:14:53
    player engagement they cite a couple of
  • 00:14:55
    Prior works like the 2017 eomm paper
  • 00:14:58
    that we talked about earlier which
  • 00:15:00
    showed that just focusing on Fair
  • 00:15:02
    matches isn't sufficient for keeping
  • 00:15:03
    players hooked this new approach doubled
  • 00:15:06
    down on engagement mix it up to maximize
  • 00:15:09
    fun and retention is the general
  • 00:15:11
    strategy one analogy that was given by a
  • 00:15:13
    net e researcher was even if you have 12
  • 00:15:16
    players of similar skill you don't
  • 00:15:19
    always make two teams of equal skill
  • 00:15:21
    maybe you know you mix High skill and
  • 00:15:24
    low skill players on each side in some
  • 00:15:27
    scenarios if that pattern leads a better
  • 00:15:29
    overw engagement but they found that a
  • 00:15:31
    bit of variety occasionally playing with
  • 00:15:34
    or against someone notably better or
  • 00:15:37
    notably worse than you could keep things
  • 00:15:40
    interesting so maybe perhaps that
  • 00:15:42
    Underdog upset or you know you face
  • 00:15:45
    someone really good and you're learning
  • 00:15:47
    from a really skilled teammate or
  • 00:15:48
    opponent maybe those things make the
  • 00:15:50
    sessions bit more memorable enjoyable so
  • 00:15:53
    rivals' matchmaking is born from this
  • 00:15:55
    lineage of research which I thought is
  • 00:15:58
    kind of cool first of all it likely
  • 00:16:00
    groups players by their General skill
  • 00:16:02
    brackets so a beginner is not going to
  • 00:16:04
    get matched directly against some top
  • 00:16:06
    tier veteran in a normal scenario that
  • 00:16:08
    just would be unfair the second thing is
  • 00:16:11
    within that broad skill group it's going
  • 00:16:13
    to consider a variety of factors so
  • 00:16:15
    things like team composition um
  • 00:16:18
    predicted fairness and probably your
  • 00:16:21
    recent performances an engagement
  • 00:16:23
    metrics too also it it might
  • 00:16:25
    occasionally and this happened to me
  • 00:16:27
    create matches where one team has a
  • 00:16:30
    carry player like a star player and the
  • 00:16:32
    other team doesn't and then it balances
  • 00:16:35
    that by giving the star player weaker
  • 00:16:37
    teammates and the other team solid
  • 00:16:40
    mid-level players and it tries to keep
  • 00:16:43
    your win rate around a pretty moderate
  • 00:16:46
    level and it's not doing that by forcing
  • 00:16:48
    outcomes I don't think it can but by
  • 00:16:51
    dynamically adjusting the challenges
  • 00:16:53
    that you face a consistently winning
  • 00:16:56
    player will be moved up against
  • 00:16:59
    opposition or maybe even given less
  • 00:17:01
    backup while a struggling player might
  • 00:17:03
    catch a break with easier opposition or
  • 00:17:05
    maybe some more support so over many
  • 00:17:07
    games this feels like some invisible
  • 00:17:10
    hand that is ensuring that most players
  • 00:17:13
    hover roughly around 50% wins so in the
  • 00:17:16
    Marvel Rivals Community a common refrain
  • 00:17:19
    that you hear is the matchmakings rigged
  • 00:17:22
    but what do players really mean by that
  • 00:17:25
    and why do they feel it is what I was
  • 00:17:27
    wondering constantly so I went out and I
  • 00:17:29
    went to forums typically from Steam and
  • 00:17:32
    Reddit and other places and and just
  • 00:17:34
    read through threads and here's here's
  • 00:17:36
    what people say the first one is that
  • 00:17:38
    50/50 win rate feeling players like you
  • 00:17:41
    and me often notice that just no matter
  • 00:17:44
    how well you play your win rate just
  • 00:17:47
    tends to gravitate around 50% so for
  • 00:17:50
    every hot streak where they just want a
  • 00:17:52
    bunch of games a cold streak follows and
  • 00:17:55
    it's this feeling that it's almost
  • 00:17:57
    looming right you're playing you're
  • 00:17:58
    killing it it's going great and you're
  • 00:18:00
    like crap some something's coming the
  • 00:18:02
    players have voiced that the system
  • 00:18:04
    seems to kind of give you wins and then
  • 00:18:07
    suddenly throw you into guaranteed
  • 00:18:08
    losses creating a really strong
  • 00:18:10
    oscillating pattern and in in different
  • 00:18:13
    forums some of the players claim that
  • 00:18:15
    the game tries to satisfy everyone with
  • 00:18:17
    a 48 to 52% win loss ratio regardless of
  • 00:18:21
    whether you're a noob or a pro it can
  • 00:18:24
    certainly feel that way and I know we've
  • 00:18:26
    all been there you dominate a few rounds
  • 00:18:28
    in a few matches and then the next
  • 00:18:30
    matches your team is just mysteriously
  • 00:18:33
    uncoordinated or the other guys are just
  • 00:18:37
    murderers so it it feels it's as if the
  • 00:18:40
    game is just a rubber band you stretch
  • 00:18:42
    too far ahead and wins and then it just
  • 00:18:44
    snaps you back with defeats from a
  • 00:18:46
    players perspective this can be really
  • 00:18:49
    infuriating like why is the game
  • 00:18:51
    punishing me for winning so in a in a
  • 00:18:53
    traditional sbmm system consistent wins
  • 00:18:56
    would simply mean that you climb to a
  • 00:18:58
    high skill bracket so you'd be facing
  • 00:19:00
    tougher opponents also ideally your
  • 00:19:02
    teammates would also be of higher skill
  • 00:19:04
    too however if Marvel Rivals is indeed
  • 00:19:07
    using engagement optimize matchmaking
  • 00:19:10
    it's not just sliding you up and down uh
  • 00:19:12
    skill ladder it might be actively
  • 00:19:14
    orchestrating the challenge level to
  • 00:19:16
    keep you in that sweet spot of
  • 00:19:18
    Engagement that it's looking for to a
  • 00:19:20
    competitor this feels like just some
  • 00:19:22
    arbitrary ceiling on their success hence
  • 00:19:25
    the feeling of a rigged system ensuring
  • 00:19:27
    you don't win too much the second one is
  • 00:19:30
    uh you know being the Team MVP on a
  • 00:19:32
    losing match this gripe is interesting
  • 00:19:34
    you know why do I always get the bad
  • 00:19:36
    teammates is a question a lot of people
  • 00:19:38
    answer well players who consistently
  • 00:19:40
    perform well say you know somebody in
  • 00:19:43
    the top 500 or someone that's just good
  • 00:19:45
    but still lose matches really are
  • 00:19:47
    starting to suspect that the matchmaking
  • 00:19:49
    is giving them weaker teammates on
  • 00:19:51
    purpose you might have a game where you
  • 00:19:54
    go 15 and two you know with cap which is
  • 00:19:58
    hard to do with cap but two of your
  • 00:20:00
    teammates just barely contribute and are
  • 00:20:02
    basically throwing it's completely
  • 00:20:04
    natural to wonder was I just placed with
  • 00:20:07
    weaker players intentionally some Rivals
  • 00:20:10
    fans believe exactly that that the game
  • 00:20:13
    knowing that you're strong is going to
  • 00:20:15
    pair you with a couple of weaker folks
  • 00:20:17
    just to balance the team's overall power
  • 00:20:19
    and maybe even you know perhaps give
  • 00:20:21
    those weaker players a chance to be
  • 00:20:22
    carried to a close match all of this
  • 00:20:25
    ties back to the research idea of mixing
  • 00:20:27
    skill levels on a team for engagement
  • 00:20:30
    while I do agree that this can create
  • 00:20:32
    some pretty exciting comeback stories
  • 00:20:34
    and it certainly has for me it does also
  • 00:20:37
    mean that as a skilled player you're
  • 00:20:40
    sometimes essentially handicapped by
  • 00:20:42
    your team because of how matchmaking
  • 00:20:44
    works and and that can definitely feel
  • 00:20:47
    like the system is rigged against you
  • 00:20:49
    one Reddit user riy noted that the
  • 00:20:52
    matchmaking creates a full profile of
  • 00:20:54
    your account and attempts to match you
  • 00:20:57
    with teammates that synergize or don't
  • 00:21:00
    based on how it wants to push your games
  • 00:21:03
    so that install loock 215 Spider-Man on
  • 00:21:06
    your team isn't just random chance in
  • 00:21:09
    other words players suspect that the
  • 00:21:11
    algorithm might sometimes deliberately
  • 00:21:14
    give you a teammate who picks a hero
  • 00:21:16
    they're terrible at so that's the
  • 00:21:18
    Spider-Man who gets two kills and dies
  • 00:21:20
    15 times in order to create a particular
  • 00:21:23
    match outcome or experience I know it it
  • 00:21:26
    sounds a little bit conspirator
  • 00:21:29
    but given the complexity of the system
  • 00:21:31
    players are just connecting the dots in
  • 00:21:34
    ways that seem plausible to them next is
  • 00:21:36
    one of my favorite terminologies the
  • 00:21:38
    losers cue and forced losses Beyond just
  • 00:21:42
    statistical balancing some players go
  • 00:21:46
    further and suggest that there's a
  • 00:21:48
    losers Quee and and that's basically a
  • 00:21:50
    hidden state where you enter after a
  • 00:21:53
    certain number of wins where the game
  • 00:21:55
    ensures you lose the next one or few so
  • 00:21:58
    they'll account anecdotes like I won
  • 00:22:00
    five games in a row and then suddenly I
  • 00:22:03
    got three matches in a row with AFK
  • 00:22:05
    teammates or complete beginners on my
  • 00:22:07
    team I could practically predict I was
  • 00:22:10
    due for a loss that's literally what he
  • 00:22:11
    said so it's really hard to prove such
  • 00:22:15
    patterns as anything more than anecdotal
  • 00:22:17
    but the perception is very powerful and
  • 00:22:20
    I I see that the the community's trust
  • 00:22:22
    is you know eroding here and there and
  • 00:22:25
    every tough loss becomes well not a
  • 00:22:27
    learning moment it should be learning
  • 00:22:28
    moment but it starts to become this
  • 00:22:30
    evidence for rigging rivals' eomm system
  • 00:22:33
    is very often accused of manipulating
  • 00:22:36
    streaks on purpose one specific player
  • 00:22:38
    described it as a psychological tug of
  • 00:22:41
    war which I thought was nicely dramatic
  • 00:22:44
    where the game just gives you the high
  • 00:22:46
    of a winning streak and then Yanks it
  • 00:22:48
    away this claim unfortunately does align
  • 00:22:51
    with what we discussed earlier to keep
  • 00:22:53
    players hooked the system might
  • 00:22:55
    orchestrate some ups and downs right
  • 00:22:57
    that's what we talked about however
  • 00:23:00
    players understandably feel that a loss
  • 00:23:02
    isn't fair if it was engineered in any
  • 00:23:05
    way if you believe that you're losing a
  • 00:23:07
    match and that was predestined by the
  • 00:23:09
    algorithm's Unseen hand that's
  • 00:23:12
    essentially the definition of rigging in
  • 00:23:14
    a competitive context and that's what
  • 00:23:16
    people think another one that I was glad
  • 00:23:19
    to see which is the first thing that
  • 00:23:21
    came to mind with all this stuff was the
  • 00:23:23
    lack of transparency or or basically the
  • 00:23:25
    blackbox effect so part of what the
  • 00:23:29
    rigged narrative here is that the
  • 00:23:31
    matchmaking system is a blackbox the
  • 00:23:33
    game in no way tells you how it decides
  • 00:23:36
    teams there's no official readout of
  • 00:23:40
    your MMR or or how it pairs you nothing
  • 00:23:43
    so when humans don't have an explanation
  • 00:23:46
    we tend to fill in the blanks if you
  • 00:23:49
    lose a match that felt unfair it's way
  • 00:23:52
    easier to blame the Unseen algorithm
  • 00:23:55
    than to chalk it up to just random
  • 00:23:57
    chance personal mistakes or just a skill
  • 00:23:59
    issue especially when you see patterns
  • 00:24:02
    over dozens of matches your brain looks
  • 00:24:04
    for a reason players also talk to each
  • 00:24:07
    other you know and I see it even in chat
  • 00:24:09
    we're sharing stories of bizarre matches
  • 00:24:11
    suspicious patterns and you know kind of
  • 00:24:14
    half-remembered quotes from Dev
  • 00:24:16
    interviews or leaked code so in Rivals
  • 00:24:18
    this case the existence of these actual
  • 00:24:21
    research papers by the developers is
  • 00:24:24
    like pouring gasoline on the fire it's
  • 00:24:27
    it's pretty rare to be honest that
  • 00:24:29
    players have such direct hints at what
  • 00:24:32
    might be going on behind the scenes I
  • 00:24:34
    know for me I've never really had access
  • 00:24:36
    to research papers that explain
  • 00:24:39
    something about an online competitive
  • 00:24:41
    game I'm playing and and that's the
  • 00:24:43
    product of this video as well so when
  • 00:24:45
    they read things or or a line in the
  • 00:24:48
    paper like fair games are not sufficient
  • 00:24:50
    to ensure player engagement from net E's
  • 00:24:54
    own paper yeah it's it's super easy to
  • 00:24:57
    jump to conclusions aha they don't care
  • 00:25:00
    about Fair matches at all they only care
  • 00:25:01
    about engagement they'll do whatever it
  • 00:25:04
    takes to make us play more even if it
  • 00:25:06
    means rigging games you bastards the
  • 00:25:09
    emotional toll is is notable I mean it
  • 00:25:12
    it feels personal sometimes games are
  • 00:25:15
    emotional experiences I would even argue
  • 00:25:17
    that Rivals is even more of an emotional
  • 00:25:19
    experience because there's a team
  • 00:25:20
    involved and this is accountability or
  • 00:25:23
    lack thereof there's this thing where
  • 00:25:26
    people are delusional and you don't want
  • 00:25:27
    to accept your own mistakes so you look
  • 00:25:29
    to others it's it's like a tiny little
  • 00:25:31
    Society it's very interesting but when
  • 00:25:33
    you lose because you got outplayed it
  • 00:25:36
    stings it sucks but you can accept it
  • 00:25:39
    and try to improve but when you lose and
  • 00:25:42
    feel helpless like nothing you could
  • 00:25:45
    have done would change the outcome then
  • 00:25:47
    it breeds frustration and conspiracy so
  • 00:25:51
    Rivals players have expressed that eomm
  • 00:25:54
    undermines the core satisfaction of
  • 00:25:56
    improving as a player and then that's a
  • 00:25:58
    quote by the way undermines the core
  • 00:26:00
    satisfaction of improving as a player
  • 00:26:02
    making success or failure feel
  • 00:26:04
    disconnected from your skill that's a
  • 00:26:05
    pretty strong psychological blow you
  • 00:26:07
    know if if practice and skill building
  • 00:26:10
    don't guarantee better results you know
  • 00:26:12
    because the matchmaking will just throw
  • 00:26:14
    a wrench in it then when Why Try it all
  • 00:26:16
    I'm spending time in AIM trainers and in
  • 00:26:18
    the practice mode what's that practice
  • 00:26:20
    mode for if I'm just going to lose
  • 00:26:22
    because the game makes it you know this
  • 00:26:23
    this perceived lack of player agency
  • 00:26:26
    drives people to conclude that the
  • 00:26:27
    system is fundamentally unfair it's it's
  • 00:26:31
    not just that one match was off it's the
  • 00:26:34
    idea that the entire ladder is a lie
  • 00:26:37
    you're you're not truly in control of
  • 00:26:38
    your rank because the game is just
  • 00:26:40
    pulling strings in the background and
  • 00:26:42
    that's even a more sensitive issue when
  • 00:26:44
    people try to rank you know when you're
  • 00:26:45
    trying to actually get better at the
  • 00:26:47
    game and earn certain rewards then it
  • 00:26:49
    starts to get a little controversial so
  • 00:26:51
    all these perceptions start to create a
  • 00:26:54
    really powerful narrative that Marvel
  • 00:26:56
    rivals' matchmaking is rigged which I'm
  • 00:26:59
    not sure if I would use that word but
  • 00:27:01
    that's what people feel it's important
  • 00:27:03
    to note here that you know not every
  • 00:27:05
    player feels this way some climb the
  • 00:27:06
    ranks and feel the only thing stopping
  • 00:27:09
    them is the next skill hurdle and that's
  • 00:27:11
    that's a fantastic mindset to have and
  • 00:27:13
    if you have that congratulations but the
  • 00:27:16
    fact that enough players share the
  • 00:27:18
    rigged perception means that it's a real
  • 00:27:21
    Community issue that I thought was worth
  • 00:27:23
    investigating which is why I'm making
  • 00:27:24
    this video now let's peek behind the
  • 00:27:27
    accusations you know how does Marvel
  • 00:27:29
    rivals' matchmaking actually work let's
  • 00:27:32
    pose the big question are these rigging
  • 00:27:35
    claims true what does the evidence say
  • 00:27:38
    about Marvel rivals' matchmaking
  • 00:27:40
    mechanics and their effects let's
  • 00:27:42
    analyze this point by Point comparing
  • 00:27:45
    player claims with what we know from
  • 00:27:47
    research and data the first claim as you
  • 00:27:50
    remember was that the system forces a
  • 00:27:52
    50% win rate for everyone so it's not
  • 00:27:56
    literally forcing a precise 50% on
  • 00:28:00
    everyone some players do maintain higher
  • 00:28:03
    win rates especially in team games where
  • 00:28:06
    you can party up with your friends and
  • 00:28:08
    communicate directly and practice
  • 00:28:10
    together or some players are just
  • 00:28:11
    amazing at their characters right and
  • 00:28:13
    they're just very talented however the
  • 00:28:16
    this engagement optimized design does
  • 00:28:19
    tend to nudge everyone towards the
  • 00:28:21
    middle so the eomm logic from the EA
  • 00:28:24
    paper and some other ones as well
  • 00:28:27
    explicitly tries to avoid long losing or
  • 00:28:30
    winning streaks because those can reduce
  • 00:28:33
    engagement if Marvel Rivals indeed uses
  • 00:28:35
    that logic then over time most players
  • 00:28:39
    will experience roughly equal wins and
  • 00:28:42
    losses by Design not because the game
  • 00:28:45
    predetermines each match outcome but
  • 00:28:47
    because if you start deviating too high
  • 00:28:50
    or too low the matchmaking adapts it's
  • 00:28:52
    just going to snap you back to the
  • 00:28:53
    middle win a lot and as a result you're
  • 00:28:56
    going to be facing tougher matchups
  • 00:28:58
    you're going to have you know stronger
  • 00:28:59
    opponents maybe even less favorable team
  • 00:29:02
    comps until you lose or lose a lot and
  • 00:29:06
    you're going to get easier matchups so
  • 00:29:08
    weaker opponents strong Ally or even
  • 00:29:11
    Bots inserted to help until you win so
  • 00:29:14
    it's it's basically a rubber band effect
  • 00:29:16
    and I talked about it earlier but racing
  • 00:29:18
    games like Mario Kart famously give the
  • 00:29:21
    leading racer weaker powerups and the
  • 00:29:23
    trailing racer powerful boost like the
  • 00:29:25
    [ __ ] blue shells just a just a keep
  • 00:29:28
    the races close and unpredictable so
  • 00:29:30
    Marvel rivals' matchmaking appears to be
  • 00:29:32
    doing a a kind of a similar thing but
  • 00:29:35
    with how it forms its matches this
  • 00:29:37
    doesn't mean that every player is
  • 00:29:39
    exactly at 50% of course we know that
  • 00:29:42
    highly skilled players can still
  • 00:29:44
    maintain a very positive win rate over
  • 00:29:46
    hundreds of games and very new players
  • 00:29:49
    might be below the 50% as they learn but
  • 00:29:52
    the system's intent is to keep you
  • 00:29:56
    around an equilibrium where where you
  • 00:29:58
    win about as much as you lose therefore
  • 00:30:01
    preserving that kind of tension and the
  • 00:30:04
    drive to play just one more match which
  • 00:30:07
    Marvel Rivals has that in Spades so
  • 00:30:09
    there there is some truth to this claim
  • 00:30:12
    but calling it forced might be too
  • 00:30:15
    strong it's it's just probabilistic
  • 00:30:18
    adjustment not a guarantee you know one
  • 00:30:21
    one Community analysis even pointed out
  • 00:30:23
    that nowhere in the available research
  • 00:30:26
    was there a statement that one team will
  • 00:30:28
    always get an advantage over the other
  • 00:30:32
    or that the outcomes are pred decided so
  • 00:30:35
    the matches themselves aren't fixed you
  • 00:30:37
    you always have a chance to win it's
  • 00:30:39
    just that the conditions are tuned to
  • 00:30:42
    make those chances chances hover over
  • 00:30:44
    the 50/50 for most players the second
  • 00:30:46
    claim was that matchmaking gives me bad
  • 00:30:48
    teammates on purpose to make me lose so
  • 00:30:51
    this one's really tricky um if we
  • 00:30:54
    interpret bad as in lower skill
  • 00:30:56
    teammates if we agree on that the
  • 00:30:58
    research does suggest that this does
  • 00:31:01
    happen occasionally by Design end match
  • 00:31:03
    findings show that a team composition of
  • 00:31:06
    mixed skill so what that means is one
  • 00:31:08
    high skill and two low skill in a 3v3
  • 00:31:11
    team for example can improve overall
  • 00:31:14
    engagement versus all players being of
  • 00:31:17
    equal skill yeah the the algorithm might
  • 00:31:20
    sometimes pair a strong player with
  • 00:31:22
    weaker ones that can happen however it's
  • 00:31:25
    not as simple as to make you lose in
  • 00:31:28
    theory those matches are still intended
  • 00:31:30
    to be winnable the other team might also
  • 00:31:33
    have a mix maybe one medium-sized anchor
  • 00:31:36
    instead of a high skilled star but all
  • 00:31:38
    pretty decent the purpose here the aim
  • 00:31:41
    is that the match is close that's it the
  • 00:31:44
    danger is that if you as a high skilled
  • 00:31:46
    player don't perform exceptionally well
  • 00:31:49
    or to to the standard that the game
  • 00:31:51
    expects you to and you might indeed lose
  • 00:31:54
    and then just feel held back by your
  • 00:31:55
    allies and what about the exam case like
  • 00:31:59
    deliberately griefing you with team
  • 00:32:01
    members who are far below your skill
  • 00:32:03
    well regular ranked play probably won't
  • 00:32:07
    put a top tier and bottom tier player
  • 00:32:09
    together often because of the initial
  • 00:32:11
    grouping by skill so probably not going
  • 00:32:13
    to happen but in agase scenarios or if
  • 00:32:17
    the population is low at your skill
  • 00:32:19
    level and that leads to to kind of wider
  • 00:32:22
    searches you might see it so also if if
  • 00:32:26
    the system thinks that you're lik ly to
  • 00:32:28
    carry and still have a fair match it
  • 00:32:31
    might gamble on a wider skill Gap team
  • 00:32:33
    so then it can happen as well but most
  • 00:32:36
    importantly there's no evidence
  • 00:32:39
    whatsoever that the system identifies a
  • 00:32:41
    specific player to punish and then
  • 00:32:44
    assigns bad teammates out of spite it's
  • 00:32:47
    not like that it's it's more systemic it
  • 00:32:50
    just treats everyone as part of the big
  • 00:32:52
    engagement equation so you might get bad
  • 00:32:56
    teammates one match and and then be the
  • 00:32:58
    bad teammate in another match and then
  • 00:33:02
    it all kind of balances out so one thing
  • 00:33:04
    that I thought really debunks the more
  • 00:33:07
    paranoid version of this claim is that
  • 00:33:10
    like the truly horrible mismatches would
  • 00:33:12
    ruin engagement not improve it if Marvel
  • 00:33:15
    Rivals put a newbie on your high ranked
  • 00:33:18
    team that newbie would be miserable
  • 00:33:20
    they're going to crush him they're going
  • 00:33:21
    to go after him he'll be angry my throw
  • 00:33:25
    you'll be angry at them that's a lose
  • 00:33:27
    lose scenario for the purpose of
  • 00:33:30
    boosting engagement and enjoyment of the
  • 00:33:31
    game in other words it would create
  • 00:33:33
    churn which is something that these
  • 00:33:35
    algorithms do take into account and try
  • 00:33:37
    to avoid at any cost so the more likely
  • 00:33:39
    scenario here I think is that you just
  • 00:33:42
    get teammates slightly below your level
  • 00:33:45
    paired against opponents slightly above
  • 00:33:47
    your team's average and then if if
  • 00:33:49
    that's done right the match could still
  • 00:33:52
    be balanced so from your POV you notice
  • 00:33:55
    that your alleys allies are making more
  • 00:33:57
    mistakes than you and it just feels
  • 00:33:59
    unfair I think that's what's going on
  • 00:34:01
    the third claim is that the game
  • 00:34:03
    secretly adjusts gameplay so damage hit
  • 00:34:07
    Boxes Etc or uses Bots to influence
  • 00:34:10
    outcomes okay this one is a bit wacky
  • 00:34:12
    but this is a deeper conspiracy floating
  • 00:34:15
    around that I've seen so some players
  • 00:34:18
    speculated some crazy things like the
  • 00:34:21
    game might dynamically Nerf your damage
  • 00:34:24
    output if you're on a winning streak or
  • 00:34:26
    expand an enemy hitbox to ensure you
  • 00:34:29
    miss essentially kind of invisible
  • 00:34:32
    handicaps Beyond just team composition
  • 00:34:34
    and one person said that that's why they
  • 00:34:36
    don't have the damage number so we can't
  • 00:34:38
    tell so as of now there's zero concrete
  • 00:34:42
    evidence for such kind know real time
  • 00:34:44
    in-match rigging that would be far more
  • 00:34:48
    akin to actual match fixing and probably
  • 00:34:52
    detectable if somebody records and
  • 00:34:54
    analyzes gameplay data I've just never
  • 00:34:56
    bothered to do that all indications from
  • 00:34:58
    the research papers that I read are that
  • 00:35:01
    the rigging if any happens is kind of at
  • 00:35:04
    the matchmaking stage not within the
  • 00:35:07
    match and I'm even careful to call it
  • 00:35:08
    rigging but you know what I mean so the
  • 00:35:10
    game isn't altering your bullets or your
  • 00:35:13
    hero's Health on the Fly under the guys
  • 00:35:16
    of matchmaking it just it would be a
  • 00:35:19
    technical and I I also think ethical can
  • 00:35:21
    of worms for the developers to do that
  • 00:35:24
    as for Bots you know marvel Rivals like
  • 00:35:27
    many games does use bots in some modes
  • 00:35:30
    or just a backfill if a human player
  • 00:35:32
    isn't available so that and then that
  • 00:35:34
    makes sense like the allegation is that
  • 00:35:37
    after a string of losses you might be
  • 00:35:39
    placed into a match where one or two of
  • 00:35:41
    your teammates are actually AI Bots
  • 00:35:43
    thisg disu this you know player like
  • 00:35:46
    names who were tuned to play well with
  • 00:35:48
    you or let you shine thereby you know
  • 00:35:51
    boosting your morale or something so in
  • 00:35:53
    community discussions some claim to have
  • 00:35:57
    noticed patterns like teammates that
  • 00:35:59
    behave oddly or too predictably and
  • 00:36:02
    suspect that they're Bots put there to
  • 00:36:04
    help them out the developers have not
  • 00:36:06
    confirmed this if true it would be an
  • 00:36:09
    extreme form of Engagement optimization
  • 00:36:12
    like literally rigging a win for you by
  • 00:36:15
    using dummy players right so it's it's
  • 00:36:17
    plausible in theory you know single
  • 00:36:19
    player games tend to adjust difficulty
  • 00:36:22
    all the time but here it would be like
  • 00:36:24
    adding an NPC on your team to just tilt
  • 00:36:27
    the odds whatever but without official
  • 00:36:29
    word or data we should just take this
  • 00:36:30
    claim with a grain of salt it's a little
  • 00:36:33
    crazy the next claim was engagement
  • 00:36:35
    optimized means that the game cares more
  • 00:36:38
    about profit than players this is more
  • 00:36:41
    of a sentiment and a feeling rather than
  • 00:36:43
    a factual claim but it it does underpin
  • 00:36:46
    the rigged narrative so players feel
  • 00:36:48
    that if the goal isn't pure competition
  • 00:36:52
    then the developers must be manipulating
  • 00:36:54
    matches to make money for example by
  • 00:36:57
    keeping you to play longer so maybe you
  • 00:37:00
    spend on a cool skin or just frustrating
  • 00:37:03
    you enough to just hey maybe if I buy a
  • 00:37:06
    new skin I'll play better I don't know
  • 00:37:07
    there's there's some truth to you know
  • 00:37:10
    that engagement often correlates with
  • 00:37:12
    revenue and freeo playay games if you
  • 00:37:14
    play longer there are more chances and
  • 00:37:16
    opportunities for you to just buy
  • 00:37:18
    something and you know we saw that in
  • 00:37:20
    the EA paper they outright mentioned
  • 00:37:23
    possibly tuning matchmaking for spending
  • 00:37:25
    metrics who knows maybe they're doing it
  • 00:37:27
    here I don't know but let's let's be
  • 00:37:29
    objective engagement doesn't only
  • 00:37:32
    benefit the company it can also benefit
  • 00:37:33
    the player too in terms of having more
  • 00:37:36
    people to play with you know a game that
  • 00:37:38
    retains players well is alive and fun
  • 00:37:41
    the community is great A game that is
  • 00:37:43
    brutally fair but Hemorrhage as newbies
  • 00:37:46
    might end up in a graveyard where only
  • 00:37:48
    the hardcore remain and I can tell you
  • 00:37:50
    for sure other games I play are games
  • 00:37:52
    like Tekken where it came off and
  • 00:37:54
    started off with a lot of players and
  • 00:37:56
    now whoever's left they just the sweat
  • 00:37:58
    Lords and it's a murder scene in there
  • 00:38:01
    so this is definitely definitely true
  • 00:38:03
    the developers likely do want you to
  • 00:38:06
    have fun just enough fun and some ups
  • 00:38:09
    and downs you know so that you can just
  • 00:38:11
    keep coming back and having different
  • 00:38:13
    emotional experiences and it's I know
  • 00:38:15
    it's it's a delicate balance if they
  • 00:38:17
    push too hard for profit and then make
  • 00:38:20
    the experience unenjoyable then it
  • 00:38:23
    backfires and players leave and we've
  • 00:38:26
    you know we've seen that many many times
  • 00:38:28
    before so while yes the system is there
  • 00:38:32
    in part to increase metrics that tie to
  • 00:38:34
    profit I don't think it's just a simple
  • 00:38:37
    Evil versus good scenario they're trying
  • 00:38:40
    to align what keeps you playing with
  • 00:38:43
    what you enjoy just with maybe a heavier
  • 00:38:46
    hand than some would like that's it and
  • 00:38:48
    they have to make money right somehow
  • 00:38:50
    and the game is free so I'd be okay with
  • 00:38:53
    a little push towards profit another
  • 00:38:55
    claim was that if I am stuck in rank
  • 00:38:58
    it's the matchmaking fault all right so
  • 00:39:00
    this is the classic ELO hell argument
  • 00:39:03
    heard in other games from you know
  • 00:39:05
    OverWatch all the way to League of
  • 00:39:07
    Legends um now eold and and Marvel
  • 00:39:10
    Rivals so the idea is that you might
  • 00:39:14
    belong in a higher rank say you believe
  • 00:39:17
    that you have the skill of a diamond
  • 00:39:18
    player but the matchmaking systems
  • 00:39:21
    tricks keep you stuck in Gold by
  • 00:39:24
    throwing unwinable games at you is it
  • 00:39:26
    true for a consistently excellent player
  • 00:39:31
    probably not in the long run there's a
  • 00:39:33
    YouTuber whose name is awward and I
  • 00:39:36
    think he's a former OverWatch guy and he
  • 00:39:39
    took mantis on 100% winning streak I
  • 00:39:41
    think he went all the way from bronze to
  • 00:39:43
    to GM over a six-hour period so
  • 00:39:46
    obviously there's outliers for these
  • 00:39:47
    things but while engagement matchmaking
  • 00:39:51
    can I mean really slow down how quickly
  • 00:39:54
    you climb because it will feed you some
  • 00:39:57
    losses even if you're good as as awkward
  • 00:40:00
    showed if you genuinely outperform your
  • 00:40:02
    Pierce significantly you're going to
  • 00:40:04
    rise so we see top players they're able
  • 00:40:08
    to reach the highest ranks in Marvel
  • 00:40:10
    Rivals we see it happen you know not
  • 00:40:12
    constantly but pretty frequently and
  • 00:40:15
    they also deal with the same system
  • 00:40:17
    right but just their skill carries them
  • 00:40:19
    through the artificial hurdles that the
  • 00:40:21
    game throws at them on the other side
  • 00:40:23
    there's one highlevel player that
  • 00:40:25
    pointed out that they've seen top tier
  • 00:40:27
    play player Smurf basically play on new
  • 00:40:29
    accounts and even they lose games in
  • 00:40:32
    bronze on occasion so that's normal
  • 00:40:35
    Varian I think over a large sample they
  • 00:40:38
    will still climb right they're great
  • 00:40:39
    players the engagement system doesn't
  • 00:40:42
    really lock doors it it perhaps adds
  • 00:40:45
    some extra zigzags to the path upward
  • 00:40:47
    but it doesn't lock doors so if someone
  • 00:40:50
    is truly hard stuck in a rank over
  • 00:40:53
    hundreds of games it's likely that their
  • 00:40:56
    current skill level
  • 00:40:57
    as harsh as that might sound and you may
  • 00:40:59
    disagree that's what's holding them back
  • 00:41:02
    the system might have given them a few
  • 00:41:04
    more bad dice rolls than a purely random
  • 00:41:07
    system would but not an impenetrable
  • 00:41:10
    wall in fact one commenter encouraged
  • 00:41:12
    that you're not in bronze silver or gold
  • 00:41:15
    because end match is rigging your games
  • 00:41:18
    you're in those ranks because you belong
  • 00:41:20
    there it's a tough love statement I know
  • 00:41:23
    but statistically it's probable the
  • 00:41:25
    engagement Focus matchmaking can impact
  • 00:41:27
    your experience of climbing so making it
  • 00:41:31
    you know Streaker or or perhaps slower
  • 00:41:34
    but there's no evidence that it stops
  • 00:41:36
    you from climbing entirely and why would
  • 00:41:39
    it so stepping back from the specific
  • 00:41:41
    claims let's weigh in what the
  • 00:41:43
    matchmaking approach does well and where
  • 00:41:46
    it falters first let's let's talk about
  • 00:41:48
    the pros um the first one that I can
  • 00:41:50
    identify as more exciting matches in
  • 00:41:53
    theory so by avoiding just complete
  • 00:41:56
    Stomps and long you know monotonic
  • 00:41:59
    streaks the system strives to give you a
  • 00:42:02
    mix of wins and losses that just keep
  • 00:42:05
    the outcome of each session uncertain
  • 00:42:07
    close matches you know come from behind
  • 00:42:10
    victories and the occasional kind of
  • 00:42:11
    nail biters are really fun I feel that
  • 00:42:14
    system on me I remember my matches I
  • 00:42:18
    remember the close calls and I remember
  • 00:42:20
    the Stomps occasionally and I remember
  • 00:42:22
    the feeling of coming back or the
  • 00:42:25
    feeling of having it and losing it
  • 00:42:26
    that's still an experience so a purely
  • 00:42:29
    Fair system can sometimes produce pretty
  • 00:42:32
    boring kind of predictable matches you
  • 00:42:35
    know imagine always knowing you'll crush
  • 00:42:37
    or be crushed with KN in between the
  • 00:42:40
    engagement approach kind of injects a
  • 00:42:43
    bit of the unexpected in there so as one
  • 00:42:45
    net e summary put it it aims for
  • 00:42:48
    patterns that improve player retention
  • 00:42:49
    by keeping things diverse another really
  • 00:42:52
    big Pro for me is retention of new or
  • 00:42:55
    casual players so new or less skilled
  • 00:42:58
    players are much less likely to be
  • 00:43:00
    thrown into a just a meat grinder of
  • 00:43:03
    constant losses I talked about Tekken
  • 00:43:05
    Tekken a before that's what it is you
  • 00:43:07
    cannot rank up these [ __ ] have
  • 00:43:09
    been playing that game for 30 years you
  • 00:43:12
    cannot it's just impossible so unless
  • 00:43:14
    you're willing to put on that work or be
  • 00:43:15
    super talented you're screwed you can't
  • 00:43:18
    if the system detects that they're at
  • 00:43:20
    risk of quitting so basically their
  • 00:43:22
    their turn probability is high then it
  • 00:43:24
    will try to give them a better
  • 00:43:26
    experience maybe that one match where
  • 00:43:29
    they get to be the hero or be really
  • 00:43:31
    impactful you know so that compassion
  • 00:43:34
    towards the lower end of the skill curve
  • 00:43:36
    typically means that more players stick
  • 00:43:38
    around which is healthy for the game
  • 00:43:41
    population and it's healthy for the game
  • 00:43:42
    and it's just no fun for anyone if only
  • 00:43:45
    the hardcore players remain and the Q
  • 00:43:48
    times Skyrocket and I guess this is also
  • 00:43:50
    why Marvel has been able to maintain the
  • 00:43:54
    player count it has pretty much since
  • 00:43:56
    launch so I see this as a really big
  • 00:43:59
    thing and I hope this never changes I
  • 00:44:00
    sort of talked about it already but
  • 00:44:02
    faster Q times is a big deal man with a
  • 00:44:05
    really flexible system like that then
  • 00:44:07
    the matchmaking can sometimes relax some
  • 00:44:11
    strict skill matching to get players
  • 00:44:13
    into the games faster if the top
  • 00:44:15
    priority was perfectly even skill then
  • 00:44:18
    higher ranked players might wait ages to
  • 00:44:20
    find the exact match rivals' approach is
  • 00:44:23
    willing to bend the rules a bit kind of
  • 00:44:25
    like mixing skill levels in a team to
  • 00:44:28
    reduce weight under the belief that
  • 00:44:30
    playing is more fun than waiting I tend
  • 00:44:33
    to agree and then the last Pro that I
  • 00:44:35
    can think of is it just encourages
  • 00:44:37
    adaptability so by sometimes putting
  • 00:44:39
    players in uncomfortable situations like
  • 00:44:42
    off rolls or just carrying weaker
  • 00:44:44
    teammates once you get good it
  • 00:44:46
    challenges them to adapt for competitive
  • 00:44:48
    folks this can actually really Foster
  • 00:44:50
    skill growth in versatility and maybe
  • 00:44:52
    even leadership you might become a
  • 00:44:54
    better player overall by learning to
  • 00:44:55
    handle a variety of M match conditions
  • 00:44:58
    not just fair fights hey these guys are
  • 00:45:00
    rushing to spawn should I just sit here
  • 00:45:03
    and chat or do I go and help you know
  • 00:45:05
    little things like that can really help
  • 00:45:07
    you out so I'm a glass half full kind of
  • 00:45:09
    person I like to think of this as a
  • 00:45:11
    positive now let's talk about the bat
  • 00:45:13
    stuff um the first one I can think of is
  • 00:45:16
    there's just this perceived lack of
  • 00:45:18
    fairness I talked about it in detail but
  • 00:45:22
    many players feel really cheated by this
  • 00:45:24
    system and I understand how that feeling
  • 00:45:27
    comes about just the the Integrity of
  • 00:45:29
    ranked competition really comes into
  • 00:45:32
    question if the rules aren't consistent
  • 00:45:35
    so in a fair match you'd say May the
  • 00:45:38
    best team win in a heavily engagement
  • 00:45:40
    driven match one could just cynically
  • 00:45:43
    say May the algorithm decide and that I
  • 00:45:46
    think undercuts the competitive Spirit
  • 00:45:49
    I'm not an amazing player I'm I'm new to
  • 00:45:51
    these games and I just hit Platinum last
  • 00:45:53
    night so I'm very proud of myself but
  • 00:45:56
    Rising platinum and now facing the
  • 00:45:59
    reality that I want to go to Diamond and
  • 00:46:00
    other ranks knowing all of this is in my
  • 00:46:03
    head definitely undercuts it a bit the
  • 00:46:06
    second one is uh a lot of high skill
  • 00:46:09
    players are just frustrated or will get
  • 00:46:11
    frustrated so the best players will
  • 00:46:14
    often want to test their skills in a
  • 00:46:16
    pure environment when they feel that the
  • 00:46:18
    system is throwing them curveballs like
  • 00:46:21
    weaker teammates or just some forced
  • 00:46:24
    weird probability scenarios it's very
  • 00:46:26
    frustrating they they don't want to they
  • 00:46:28
    don't want training wheels or artificial
  • 00:46:30
    difficulty they want their true skill to
  • 00:46:32
    be the only factor and I get that some
  • 00:46:35
    are just going to quit out of
  • 00:46:36
    frustrations or or a sense that the game
  • 00:46:38
    isn't rewarding their time and we see
  • 00:46:41
    this sentiment in in the Marvel Rivals
  • 00:46:43
    Community a lot with some skilled
  • 00:46:44
    players announcing their departure due
  • 00:46:46
    to the eomm system then then I can also
  • 00:46:48
    think of just it's complex and it's
  • 00:46:51
    unpredictable so the system is so
  • 00:46:54
    complex that even develop
  • 00:46:57
    and data scientists struggle to
  • 00:46:59
    perfectly tune it and that tells you
  • 00:47:01
    something that means it might make some
  • 00:47:04
    mistakes a lot of mistakes so you could
  • 00:47:06
    get matches that end up just horribly
  • 00:47:09
    one-sided or bizarre outliers that's
  • 00:47:12
    slip through the algorithms you know
  • 00:47:14
    testing and when it does players won't
  • 00:47:17
    just see it as a one-off error they'll
  • 00:47:20
    potentially view it as proof of
  • 00:47:22
    malicious rigging and then they can get
  • 00:47:24
    really pissed so the lack of
  • 00:47:26
    transparency amplifies this players just
  • 00:47:28
    don't know why a match was bad so they
  • 00:47:31
    just assume the worst the the last con
  • 00:47:33
    I'll give you here is just the ethical
  • 00:47:35
    and long-term issues it proposes so if
  • 00:47:38
    players feel manipulated that erodes
  • 00:47:40
    trust in in the short term they might
  • 00:47:43
    keep playing the the Skinner box effect
  • 00:47:46
    if anybody knows it or of variable
  • 00:47:48
    rewards they can be pretty compelling
  • 00:47:50
    sometimes but in the long run it can
  • 00:47:52
    foster a lot of cynicism and burnout so
  • 00:47:55
    now we're in the honeymoon period period
  • 00:47:57
    I don't know what it's going to look
  • 00:47:58
    like in 6 months so you know a game can
  • 00:48:00
    develop a reputation for being unfair or
  • 00:48:02
    pay to win in some cases even if money
  • 00:48:04
    isn't directly involved kind of any
  • 00:48:06
    width of trying to get engagement to
  • 00:48:08
    monetization can really tarnish
  • 00:48:10
    perception so if if Marvel Rivals is
  • 00:48:13
    seen as too psychologically manipulative
  • 00:48:16
    it might drive away the exact dedicated
  • 00:48:19
    community that would sustain it for
  • 00:48:20
    years and I I know you know I I've made
  • 00:48:23
    Marvel snap videos Marvel snap is a
  • 00:48:25
    Mobile card game that suffers from
  • 00:48:27
    exactly that and if you want to kind of
  • 00:48:29
    see a Telltale case of that and people
  • 00:48:31
    leaving and leaving the game behind me
  • 00:48:33
    included this is why I'm making Rivals
  • 00:48:35
    content now and I don't cover snap
  • 00:48:36
    anymore this is it right there oh hello
  • 00:48:39
    person that skipped the whole video just
  • 00:48:40
    to get to this
  • 00:48:42
    part anyway so what is the conclusion
  • 00:48:47
    rigged or just reality well the answer
  • 00:48:50
    is both simpler and more complicated
  • 00:48:53
    than the question implies it's rigged in
  • 00:48:56
    the same same way that a theme park ride
  • 00:48:59
    is rigged the ups and downs and thrills
  • 00:49:02
    of it are engineered they're planned
  • 00:49:05
    it's just a bunch of metal but the fun
  • 00:49:07
    you're feeling and the puke that's
  • 00:49:09
    coming out of your mouth that's entirely
  • 00:49:11
    real so in a just purely competitive
  • 00:49:14
    sense rigging suggests deceit so it's
  • 00:49:19
    the idea that the outcomes are
  • 00:49:20
    pre-ordained or that someone is pulling
  • 00:49:23
    strings so that the matches aren't on
  • 00:49:25
    the same level Marvel rivals'
  • 00:49:28
    matchmaking isn't rigged in that overly
  • 00:49:31
    nefarious way no one is sitting at a
  • 00:49:34
    control panel somewhere saying give Nerf
  • 00:49:37
    pool a loss he's been kicking ass that's
  • 00:49:40
    not happening there's no evidence of the
  • 00:49:42
    game flat out choosing winners or
  • 00:49:45
    tweaking your damage in real time to
  • 00:49:47
    make you lose a fight come on like
  • 00:49:49
    that's nuts you can always through skill
  • 00:49:52
    and teamwork locking in good
  • 00:49:55
    communication just overcome the
  • 00:49:57
    challenges that the system throws at you
  • 00:49:59
    nothing is unwinable or unwinable by
  • 00:50:02
    Design in absolute terms however Marvel
  • 00:50:06
    Rivals is for sure using a sophisticated
  • 00:50:09
    matchmaking algorithm that has its own
  • 00:50:11
    agenda that is proven in the papers and
  • 00:50:15
    also through the analysis that I've made
  • 00:50:16
    in this video so the agenda isn't purely
  • 00:50:20
    to create Fair even contests it's just
  • 00:50:23
    to keep you playing and enjoying or or
  • 00:50:26
    at least being engage with the game
  • 00:50:27
    that's it and yeah in the pursuit of
  • 00:50:30
    that it may at times feel like an
  • 00:50:32
    invisible hand is orchestrating your
  • 00:50:34
    rise and fall you know the system giveth
  • 00:50:37
    and the system taketh away and it
  • 00:50:39
    doesn't even tell you when and it
  • 00:50:40
    doesn't tell you why so perhaps the
  • 00:50:43
    truth is best seen as just a nuanced
  • 00:50:46
    explanation rather than just a binary
  • 00:50:49
    verdict so rivals' matchmaking lies in a
  • 00:50:51
    gray Zone and and here's why it strives
  • 00:50:54
    for balance first of all but not the
  • 00:50:56
    traditional kind of balance we expect in
  • 00:50:58
    competitive rank Play It's just
  • 00:51:00
    balancing engagement metrics with
  • 00:51:02
    fairness also it protects new or
  • 00:51:04
    faltering players which is arguably a
  • 00:51:07
    really good thing but in doing so it it
  • 00:51:09
    might hold back really good players
  • 00:51:12
    momentarily just to level things out it
  • 00:51:14
    also doesn't want you to leave angry or
  • 00:51:17
    bored but ironically that very effort
  • 00:51:21
    can make people angry if they become
  • 00:51:23
    aware of it so that's kind of silly for
  • 00:51:26
    casual players this system can be really
  • 00:51:29
    a blessing it just Smooths out the
  • 00:51:31
    extremes and ensures that more games
  • 00:51:34
    feel competitive for competitive purists
  • 00:51:36
    it can feel like a curse so it's just
  • 00:51:39
    this everpresent handicap system that
  • 00:51:41
    that kind of muddles the true display of
  • 00:51:43
    skill so in my investigative Journey
  • 00:51:47
    here I saw that the idea of Engagement
  • 00:51:51
    based matchmaking didn't just start with
  • 00:51:53
    Marvel Rivals it's it's part of a
  • 00:51:55
    broader evolution tion in gaming just
  • 00:51:57
    from the early ELO days of Chess to
  • 00:52:01
    today's kind of machine learning driven
  • 00:52:03
    techniques Marv Rivals is just one of
  • 00:52:06
    the higher profile cases where the
  • 00:52:09
    community spotted the man behind the
  • 00:52:10
    curtain so to speak thanks in part to
  • 00:52:13
    those research papers like the eomm and
  • 00:52:15
    opt match knowing about them has kind of
  • 00:52:18
    ironically made players really
  • 00:52:21
    suspicious transparency is a
  • 00:52:24
    double-edged swordman if players don't
  • 00:52:26
    like what they see they start to lose
  • 00:52:27
    trust Marvel rivals' matchmaking is just
  • 00:52:30
    engineered in a way you know not to
  • 00:52:33
    guarantee specific outcomes but just a
  • 00:52:35
    shape an overall experience and whether
  • 00:52:38
    that crosses the line into feeling
  • 00:52:41
    unfair is really a subjective topic and
  • 00:52:45
    it really varies player to player so in
  • 00:52:47
    the end we we just must reflect on what
  • 00:52:50
    we want from our games if the answer is
  • 00:52:54
    pure competition and merit-based
  • 00:52:55
    progression
  • 00:52:57
    then Marvel rivals' system is just going
  • 00:52:59
    to feel antagonistic to that maybe even
  • 00:53:02
    unacceptable for a lot of players but if
  • 00:53:05
    the answer is exciting matches and a
  • 00:53:08
    healthy game Community then you know the
  • 00:53:10
    systems goals start to make a lot of
  • 00:53:12
    sense even if the execution might need
  • 00:53:15
    some tuning one thing is very clear to
  • 00:53:17
    me is that the discussion Marvel Rivals
  • 00:53:19
    has sparked will be important for the
  • 00:53:21
    future of online games it it forces
  • 00:53:24
    developers to consider how much much
  • 00:53:26
    they should prioritize engagement over
  • 00:53:29
    transparency and fairness and it does
  • 00:53:31
    push players to ask how much algorithmic
  • 00:53:35
    meddling they're comfortable with and
  • 00:53:37
    that's going to become more and more of
  • 00:53:39
    a thing especially with AI on the rise
  • 00:53:41
    so we almost have to get used to it and
  • 00:53:44
    as players I think being informed is our
  • 00:53:47
    best tool understanding the forces at
  • 00:53:49
    play shows that we can adapt and just
  • 00:53:52
    make our own choices to play to pause or
  • 00:53:55
    just to move to to a different game if
  • 00:53:57
    that aligns with our philosophy and that
  • 00:53:59
    was the purpose of this entire
  • 00:54:01
    monstrosity of a video is to inform to
  • 00:54:04
    help you see different perspectives of
  • 00:54:06
    of this matchmaking to also nerd out a
  • 00:54:09
    little bit I think it's really cool that
  • 00:54:11
    things have evolved to to this extent
  • 00:54:13
    all the way from chess now to here and
  • 00:54:16
    I'm excited to see where it goes because
  • 00:54:18
    I do like that it's focused on shaping
  • 00:54:21
    an experience that the developers intend
  • 00:54:24
    I just have some combating feelings with
  • 00:54:28
    how healthy that is for a competitive
  • 00:54:30
    environment but overall it was
  • 00:54:32
    incredibly fun for me to learn all of
  • 00:54:35
    this it was exhausting to make this and
  • 00:54:38
    to research everything so hopefully you
  • 00:54:40
    appreciated it and um I'm I'm looking
  • 00:54:43
    forward to seeing how net e you know
  • 00:54:46
    advances with this and even though I am
  • 00:54:49
    don't want to look at a research paper
  • 00:54:51
    again in the near future I would be
  • 00:54:53
    excited to see new developments around
  • 00:54:55
    this topic cuz I think it's genuinely
  • 00:54:56
    kind of cool so the next time you find
  • 00:54:59
    yourself in a baffling losing streak or
  • 00:55:03
    an effortless win streak in Marvel
  • 00:55:05
    Rivals you'll know that it might not be
  • 00:55:07
    mere luck of the draw there's a system
  • 00:55:10
    in the shadows a balancing and
  • 00:55:13
    rebalancing of the scales rigged not
  • 00:55:16
    exactly but neither is it the Level
  • 00:55:19
    Playing Field we once imagined it's
  • 00:55:21
    something new a Battleground where human
  • 00:55:25
    skill and machine strategy
  • 00:55:28
    intersect and in this new Arena the
  • 00:55:31
    final judgment on whether that's right
  • 00:55:33
    or wrong is up to each of you
タグ
  • Matchmaking
  • Marvel Rivals
  • EOMM
  • SBMM
  • Online Gaming
  • Player Engagement
  • Game Design
  • Skill Rating
  • Player Experience
  • Competitive Gaming