Design Club: Why We Need Skill Based matchmaking

[music] [music] Good time zone, internet. Hello and welcome to Design Club, the only game design show on the internet that starts late today because I had another thing. And Jeff, let me move the show. I am your humble host, Will, joined as always by my pal Jeff. Say hi to the people, Jeff.
Hi to the people, Jeff. Hey everybody, I am Jeffrey Zack in the studio director of Extra Credits and Extra History and more importantly for this longtime game designer joined as always by my co-host Will, Voice of the People.
Hello.
Say hello, Will.
Hello, Will. See, I can do the joke, too.
So, uh,
you left out a couple words, but yeah.
Yeah. No, it's fair. Uh, I'm I'm I am continually infuriated by YouTube, which is why I'm leaving out words. Uh, I turned off the new autocrop us a vertical because it doesn't work because it crops us right here in the middle of the screen where there's nothing because there's two people and we need balance

in the screen. So, the people who are watching the vertical are just getting an empty segment and I can't turn it off without completely breaking the stream. So, that's fun. I'll worry about it later, but it's annoying me. That that is all. Uh,
hey Will, in my piece of fun completely irrelevant but fun news. Yes.
I have a pocketk knife
that I've had for a decade.
Uhhuh.
That's a tiny little fold out like katana
that I finally lost going through airport security on the way home from packs.
Yeah.
Like a decade in cuz I forgot it in my pocket.
Uhhuh.
And I found that the company that I kickstarted it with
10 years ago actually now just sells them. So if you've never mine my last one was red, but see this tiny little knife?
Yeah.
It's a fold out.
But this is my
Yeah. This is my brand new tiny little pocket katana.
Yeah, very nice.
Because a little pocket knife is a useful thing.
Absolutely. Uh I I have used one uh regularly and usually carry one with me

for a little of everything.
I had the option uh Thunder to go back through all the way security, find a way to mail it myself and then get back in the line for security and miss my airplane or surrender the knife. And at the end of packs, that many days in, I was just
here. You can just have it.
Time to find a new one.
Yeah, absolutely. Um,
a decade.
A decade. Yeah, that's that's a pretty good track record.
So, today we are talking about skill-based matchmaking.
Huzzah. And we want to lead off with the why we're talking about skill-based matchmaking because it there's a big discussion to have about this, right? because it is a big complicated topic with a bunch of little facets about player experience and game design and like can your game support skill-based matchmaking and what the axes of evaluation are.
Oh, and actually quick note um we have done an episode on this before and I'll drive Will crazy by saying you can see it right here because Will isn't able to

do a popup on these and so it's actually kind of funny actually. Wait, we could there's a way. No, anyway, you can see it. I'll I have to do it after the thing now. So when it'll be there right now,
which is talking about the mechanics of how you build like an MMR or skill-based matchmaking system and the systems behind it. That's actually not what we're talking about today, but this is the skill-based matchmaking on why do we have it and um is it good for games? like more of the philosophical skill-based matchmaking versus hey, how do you build a better skill-based matchmaking system that works great for games?
Yep.
Uh okay. So,
and um as actually the TLDDR from that episode was designing matchmaking is not easy, but it can be done well.
Correct. Yeah, absolutely. So, we're talking about this right now because the

there's a a a petition a petition that as of recording has hold please 16,410 verified signatures was covered by IGN where I saw it that petitions um the world I guess
but specifically developers of Call of Duty to remove skillbased matchmaking from Call of Duty. And I'll I'll read I will read the the issue to you from that.
Please do.
Call of Duty public multiplayer needs to feel casual again. For many of us, the excitement of classic Call of Duty came from the unpredictability. You could dominate one match, struggle the next, beat all kinds of players along the way. That variety made the experience memorable and gave you a reason to keep playing. Former Call of Duty developer Brian Bright recently shared that his team's work on NX1 focused on ping and region. He also stated that Activision

eventually began hiring data scientists to work with its marketing department to push for more skill-based matchmaking. His statements raise serious questions about how matchmaking priorities evolved. For players like us, today's public matches too often feel like the same competitive experience on repeat. Experimenting with a different weapon, relaxing after work, or playing with friends of different skill levels can become frustrating. When every lobby feels like another test, the fun starts to disappear. We are asking Activision and Call of Duty's development teams to remove skill-based matchmaking from standard unranked public playlist and prioritize connection quality. Keep skill-based competition and ranked play where players deliberately choose that experience. We also want clear communication how matchmaking works and meaningful opportunities for the community to test alternatives. Call of Duty became part of our lives because it was fun, social, and unpredictable. It's time to bring that feeling back. So, and uh uh some someone pointed out quite correctly Odin here quite correctly pointed out that this is a um this is a relatively small sample size. it is, but

it shows that there are a number of people who hold this viewpoint. And I want to talk about it because this is a fascinating concept to me because it's not the first time that I've heard um opposition to skill-based matchmaking specifically in public casual shooter lobbies, right? Um as a a talking point, right? This is a thing that comes up a non-trivial amount of the time, pretty much every time there's a new Call of Duty, they're like, "Will it or won't it have skill-based matchmaking, right?" And that's like a selling point for either side when they find out what it is. Some people do like it, some people don't. Some people are fine with it and don't care. I would wager that the majority of people don't care enough to like actually weigh in and that there are some people who are very pro and some people who are very con. Um, Yeah. And I'm going to reput the one I put from grass

click.
Yeah.
Um grass said, "I would bet a decent sum of money that the people asking for this do not realize what they're asking for." Yeah.
As they would likely get stomped as well.
I agree.
And I think that's right.
Before I get into that, the letter it mentioned two interesting things. Number one is that some forms of matchmaking, like skill-based or not, one of the things they're looking for is region and ping, which is they're trying to make sure that all of the players have a roughly similar connection when they do matchmaking. So that because it can be really frustrating that if you are if you are playing against people with very low latency and you say have high latency that it feels like you're moving a little bit sluggishly and they're not and that just leads to a bad experience. So matchmaking for roughly trying to get everybody a similar style of connection is a thing you pretty much always want.
Um because also that can make otherwise

your games can also feel very variable.
Correct. You might feel fine one game, sluggish the next, you know, super sluggish the one after, and it makes it feel weird. The other of which is mentioning wanting to play with friends. This makes matchmaking really, really hard for a number of reasons because a it should be done. If Will and I want to play together, that's fine. Even though I'm on the West Coast and he's on the East Coast,
right?
But we start to get remember that thing I just mentioned about ping and latency. So you're not only juggling will and my skill level hopping in, but also our connections for being able to play into the same place from all the way across the country. Let alone if we want to say, "Hey, let's have David jump in as well, one of our artists who's in Spain and maybe Rob join us as well as in Hong Kong." Like some of these games get really tricky when you want to let people play with friends.
And that's one reason you see things like League of Legends where they have regional servers. So, you know, if

you're playing on the North American servers, sure, Rob and David could hop in, but they'd have to make accounts on the North American servers and specifically say, "We're signing up for a worse experience
so that we can play with our friends."
And it's also kind of interesting because that means that you have a separate skill rating on the American servers, which means that your skill rating is different for different ping levels, which is really smart because not everyone plays the same with ping. Um, this, by the way, affects more than just competitive uh multiplayer, right? This affects every game that uses the internet. So, think about um
catching stuff in peak or uh playing an MMO where you're trying to align buffs, right? you're trying to [clears throat] line everyone's buffs, line everyone's buffs up to provide maximum effect or um you know
or you're playing Uno and you don't want the half second lag yelling uno
to try when the person plays their last card like even really simple things.

Yes,
timing can make a big difference.
Right? So that is one axis of matchmaking, but it's and it's an important one and it's a very technical one and it's not what we're focusing on.
Not what we're here to talk about today,
but it is worth acknled.
The friend part though is interesting because the friends will carry into some of the skill-based stuff.
Absolutely. Um
but anyway, just because that was in the the letter like that was an interesting thing to address.
Yep. Absolutely. Okay. Um,
so Will, the big question
I want to jump into this and talk philosophically for a moment.
Sure.
Which is why do we have skill-based matchmaking? And I wrote a whole small paragraph of stuff.
Do you want me to jump in? Do you want to jump in with something first?
Uh, I I'll give you my short answer and then and then you can expound on it maybe. I like skill-based matchmaking because I want every game to be a game in which I

am challenged
and have the possibility of improving. Get this. Regardless of whether I win or lose,
regardless of whether I'm playing casually or competitively, I like playing games on even footing so that I can improve as a player and I can feel like I've actually won something, right? Seal clubbing is not fun for anyone.
Put an asterk on that
because I think it actually is for a small number of people, some of which are the ones who wrote this pet or who signed this petition
and I'll get to that.
Okay, cool. Great. Jeeoff, why do what's your give us your give us your your thesis? So number one um is that let me actually start with this um people like to win
like straight up winning is fun and

certain games and I'm going to come more to this later. The winning is more of an essential part of the game than in other games.
Sure. And with games where the win is a majority of the part of it, and again, I'm gonna come back to this because this is also a fascinating subtopic, it's it feels weird to say, "Hey, I'm better than 70% of the people out there, but I'm only winning half my games." So, I have a young child who's in elementary school right now, and he and I play Mario Kart, and he and I play Pokémon Go, and he likes battling other real people. And he's good enough, like at Mario Kart, that he's racing against like eight people in the 8,000 rank range. He's actually better than I am now, which is for how young he is. and he likes to win, but it's tricky in that

game because they match against a lot of people, but he's getting put against better and better people and that's a little disappointing to him. But also, there is no ability of infinite scaling like on that. But more importantly, in Pokemon Go, which is a head-to-head game, again, he likes to win. Young kids like to win. As you get older, I think you start liking games that are fair and matched, balanced, but there's a certain group of people who never grow out of the I just want to win all the time,
right?
But going along with that, people like to feel like they have a chance in a game. And you know, people don't like to be pub stomped. And it's very unhealthy for a game, for a multiplayer game to have large quantities of people who are having bad match experiences all the time.
Yes.
And so part of what you're doing as the people designing the game is you're looking at your community and you're

saying, "How can I give the vast majority of the people the best experience knowing that we can't make it perfect for everybody?" Because if you let everybody play anywhere they wanted with no matchmaking, there will be certain groups of people who get to be very good and all they want to do is play in the areas with the people who aren't very good so that they can go in and win all the time. But the problem is the five or 10% of the people doing that are going to make 80 to 90% of the people they're playing against feel awful all the time because they're just going to come in and they're going to smash every time. You know, if we were talking, say Slay the Spire was a competitive game and you got you Jorbs just hung out in the low ranked massacring people or you know if League of Legends didn't skill rank and you got Faker and Double Lift and a couple of people
hanging out in the new players area

slaughtering them as they're learning the game. Like
yeah,
you say, "Oh, this isn't what we're talking about." But it absolutely is. certain people behave that way and if you don't start matchmaking them and we can see it because you get competitive games where people will continually go and make new accounts specifically so that they can go in and they can basically smurf all the time, right?
And just go stomp people.
Smurfing is a technical definition real quick is the process of making an alternate account disconnected from your main account often against terms of service though not always. um for the sole purpose of disguising your skill rating so that you play against people who are not as skilled as you. [clears throat]
That's
and like we've seen this forever.
There used to be um 30 plus years ago near my house some there was a group that actually had set up a live battlete game. I forget what it was called, but

you actually sat in pods.
Mhm. and you went and and I remember one time I played a couple going in for like a birthday or something and I'd played probably five or six matches and I was playing against a whole bunch of people that had never played
and my mech ran riot through them just 100% to the point where the person afterwards was like you've played a lot before and I had to kind of like downplay that because I I realized I ruined a whole the other team's experience because I just ran through this.
Yeah. Or, you know, when I started playing PvP and Pokemon Go, I had a 15game win streak that was just stomping people because I knew how up until my matchmaking kicked in because I knew how to fight,
right?
And build a team in this and other people obviously didn't,
right?
And it felt bad for them and you could tell it feels bad for them.
Yes. And again, as game designers, we're trying to get the best situation for the most people. And people like to feel like the game was close. And also, a blowout in either way can feel like a

waste of time. That if you're a bad player getting stomped, the game feels long and like you didn't have a purpose or a point. But also, if you're super high and, you know, slaughtering noobs isn't your thing. You're in for, you know, a good match and competitive. Having a really weak team can also feel like a waste of time because people don't have unlimited time to play. You know, you get home from work, you have time to go play two or three matches of whatever it is, and if one of them is just a blowout, you're like, man, I could have had a much better match in that limited amount of time window that I got to play this.
Yeah. And that's and that um as an example of the the match quality uh win or lose kind of sucking for everyone involved,
think about MOBAs when someone leaves. When someone leaves a mobile on either team, the whole game goes to crap, right? It's like you It's entirely It's difficult, but it's possible to win five on six, right? It It is possible [clears throat] to do that
or five on four depending on your MOA.
Yeah. Or five on four depending on your

mobile. Sorry, I'm picky deadlock. Um but the thing that is most disappointing is when it happens like 20 minutes into the game and then everyone just kind of stops caring and the quality of the match just dissolves into crap, right? because one team feels like they don't have they don't they can't win and the other team feels like it's inevitable that they win. So both teams start playing sloppy and the match degrades to the point where like it's not even really worth playing but you have to hang out until the match is safe to abandon or you get the abandoned penalty too and like now you're just like wow this just sucks. Just sucks.
Yeah. And it is like again a blow up can feel like a waste of time and it is like children want to win all the time but at a certain point playing games you get past the you know you're a six or a seven year old or whatever it is and you're actually playing and it's also it's hard to get better playing against people when you're vastly different skill levels. If you're slaughtering noobs you're not going to get better at that point as a player. If you're matched against people radically harder

than you are, you're not going to have the time to develop the chops to get better. You need people, they can be better than you, but they can't be better than you, right?
Or they can be a little lower than you, but they can't be, you know, on the floor.
Yeah. And um and and this is a this is a great example of this. Um the uh Dead by Daylight recently did a MMR rework that I've heard a lot about because some a lot of the friends I hang out with in voice chats talk about this because they play a lot of DVD. My partner plays a lot of DVD. I've heard a lot about this. They completely reset everyone's MMR to baseline in order to redo the algorithm, which had the terribly unfortunate result of people who had installed the game for the first time last week paying playing against content creators who have made Dead by Daylight their primary game and their primary career for the past multiple years.
And it's already an asymmetrical game, right? So, it's a fouron-one

asymmetrical sort of time management killer versus survivors kind of game. The intricacies are not terribly important, but suffice to say, if the one person is a person with 10,000 hours of the game and the four people all installed as a group of four friends to play yesterday, it's not going to go well, right? And that that is causing friction and tension in the community because of how incredibly unbalanced it is for both sides, right?
Yeah.
Yeah.
And also for for reals, people like a measure of how good they are
versus the rest of people. And if you're without some sort of ranking or matchmaking system, it's hard to know how good you actually are. Yeah.
You know, if you're like, "Oh, yeah. I win all the time cuz I play in public servers with no matchmaking,
right?
Like are you good? Um,

and as I said, kind of the other part is you don't want to give a place for the trolls to live purposefully live. And there are people, a small number, but significant enough that get their enjoyment out of causing other people distress in games that if you purposefully give them a place where they can go do that, if where they can go actively troll that they will do so.
Yeah. And by removing matchmaking, removing the ability to sort people into their skill level, you are setting up situations where those people can purposefully go troll by smurfing and just go and wreck other people's days.
Yeah.
And there are people that really do get enjoyment out of that.
Yeah. I'm not saying anybody in chat, but like I've worked on enough big multiplayer games to know that that is a sizable enough portion of the population that if you don't design for them and you have to design for them in a

meaningful way. Like this isn't, oh, I'm walking away from my game or you know, oh, I'm using language or other things. This is I'm just going in a place where I'm better. And the way you that you solve or you disincentivize or you stop the trolling in that place is you make a way that they can't go play versus lower level players, which is skill-based matchmaking.
Yep.
So, that was kind of my litany of like initial just gush I wanted to get out on. There's a lot of reasons that people don't necessarily think about
Yeah.
that game designers have to because they want the healthy community.
Yeah. I yeah, I think that that's a very good overview of of the the driving force behind this. Um,
yeah, I think that that's
but that other side of, oh, we want it to feel casual, like I understand that,
but there is no way to do casual that doesn't also allow for the downsides, right,
of

troll/murf behavior of sitting there and slaughtering the noobs.
Yeah.
Or even the mid levels. Like I play League of Legends. I'm roughly gold ranked. I don't want diamond level players or, you know, mast's level players coming in and just blowing out the game.
Yeah. And it's u it's also difficult because um there's a conversation going on in chat and I'll I'll relate it to a a different game because I think it'll be a little easier to see. Um in Magic Arena, you have two ratings, one of which is known and one of which is unknown. And then in some games you have a third factor in your matchmaking. You have a rank which is um based on a number of points that you get. You get one point for every game you win. You lose one point for every game you lose. Pretty simple. Past gold uh which is very early on. So you need to reach a total of plus however many to make it to

um the top rank. And you have a floor at the uh at the base of each rank as you go. So you can't once you're in platinum, you can't drop below platinum four. You'll always floor out at platinum.
That becomes your floor
and then you have to climb up. Okay, great. Awesome. So that's one of them.
But astute viewers may notice that you can climb in that system whether you have a 51% win rate and are playing 10,000 games or a 75% win rate and playing a handful because all it cares about is the difference. like you only need to reach a certain number of wins versus losses, right? And 51% will get you there eventually. I have in fact talked previously when playing games about like in 1v one ladder climbs, it is often advantageous to play the fastest deck with a positive win rate.
That is to say a win rate above 50%. In order to get there, right? And then second, you have a skill-based matchmaking rating which is hidden. And

that is the other that is the thing that sometimes confuses people is that they will say oh skill b skill-based matchmaking is not occurring in all games like arena because I a platinum player have been have been paired up against a silver player for this game and there are a whole host of reasons why that might be um some of which are not obvious to the player and that that I think is one of the reasons that this will um that this issue comes up is that It's not clear to the player what factors are actually going into them being matched. So, it comes across as skill-based matchmaking is arbitrary. I saw someone in chat saying that skillbased matchmaking was just a way to feed people the whales. Um, I pretty strongly disagree with that, but we don't have time for that conversation right now. Um,
and so I think that that is another cause of frustration in this case is not knowing why or how the skill-based matchmaking works, which is why that

part was in the letter as well, asking for increased transparency.
And it's tricky, too, because um like skill-based matchmaking only applies to competitive multiplayer games. Like if I'm playing Slay the Spire, there isn't skill-based matchmaking because I'm playing and I might post my win rate against somebody or there might be a leaderboard.
Yeah.
But that's a very different style of thing because then you're competing against a leaderboard or with your friends.
Um and when you think about it, there's kind of like three kinds of matchmaking to talk about in balance.
Yep. Um because there's head-to-head which you balance for individual play. Like if we're hopping in Magic Arena or Clash Royale or fighting games, if I'm going to go play Street Fighter or Mortal Kombat, you know, it's looking to match me as a player against another

player of roughly my same ranks skill level slash account skill level,
right? And that's I would say the easiest one to balance because there are the fewest confounding factors. It is you,
right,
versus someone else,
right?
One v one.
Because as soon as you go to the next group, which is team versus team, where you're balancing teams, like your Call of Duties, your League of Legends, your tactical shooters, Siege, Valerant, Counter Strike, it's not just, you know, Will's level versus mine. It's our team and, you know, have I brought in a friend? Okay. How do you balance it when one friend is this rank and the other friend is this rank? And
do you do sum of MMR? Like because that's that there is an argument that that is the correct way to do it, right? There there's an argument that it's sum of whatever statistical matchmaking ranking. So, it's sum versus sum and the two teams sum to the same thing. There's

another argument [clears throat] that says it should be the average, right?
Mhm. Um,
and it's different and it and honestly there probably isn't a single correct answer because what might work for say League might be different than Counter Strike.
Yeah. And like Yes, 100%. Because someone's ability
so headshot and Counter Strike is probably more valuable correct
than the slightly quicker reflexes in League for how much one person can carry.
Correct.
Versus not.
Right. Because because MOBAs are econ, there is a certain amount of
um friction against individual people hard carrying an entire team. Uh though MOA players will say that this has occurred, they mean it within the context of a MOA. It is very different than one person's mechanical skill like Shroud playing Valerant, right? Like cool. It doesn't really matter who the

other four people on that team are. One of them is Shroud. Right.
Good luck.
Yeah. Where, you know, unless it's faker, you know, League of Legends, you know, I've seen all the time top lane loses hard in League of Legends. So, like the other side is, you know, carrying hard on that. But because mid and bot did well,
right,
even though that one player up there, like they didn't get enough econom. So, they they balance out in different ways. Yeah. But balancing teams is considerably harder and that's usually the public lobby right
kind of thing which is you're you know roughly trying to figure that out. So we're not saying that there can't be improvement on it but like it's a very different style of thing. And then there's also like the free-for-all or tier ranked finish style game which is you don't win most of the time like Mario Kart. Hey, I'm fighting against or racing against 24 other people or Fortnite battle royale games where it's

like, okay, here's your group of people. Thud.
Yeah. And u Apex, I've been watching Apex as like my side monitor content in the evenings. And Apex has like this fascinating dynamic with ranked play where like they've gotten to the point where people are aware that they're not going to win, so they're just trying to get to positive ranking points. And it's such a big deal that you can see what change you your ranking point your your RP will have at the end of the match during the match.
Oh, interesting.
Everything impacts that RP and it does it live
because that it is so difficult to put together an Apex lobby that doesn't have top tier players and people climbing towards top tier that like and they tell you what the distribution is as you load in. But this that kind of like free-for-all or like multiple team or Mario Kart or whatever is unbelievably difficult to balance. Well,

um because like so many factors, even if we do Mario Kart and whoever gets whoever starts with the highest matchmaking rating is in last, that means that whoever has the highest matchmaking rating is now most likely to get a blue shell. What what I've actually noticed from playing a lot of this recently
Yeah.
is that um they have an asymmetrical point gain and loss. Yeah.
Like if there's a couple of 8,000 ranked players in mixed in with the 5,000s and 3000s getting fifth place as an 8,000, you might still lose points. We're getting 10th or 11th place as a 30,000, you might go up like 50 points.
Yeah. like they they disproportionately do that so that you feel even if you finish mid as a lower one that you did well and you're getting rewarded.
Yeah. Although that does run again without um harping too much on it that does run into the transparency problem which is one of the things that I think leads to this dissatisfaction which is

right
cool I see that that has occurred but why though
right but why though is like a really big question for this because especially if you're going to spend enough time in a system to get good right like one of my one of my biggest beefs with team fight tactics because I tried it out and team fight tactics ICS is or was when I tried it pretty opaque and the degree of success or failure was like not clear to me as a new player. Um at least in game. I probably could have gone and like done a bunch of digging to like figure it out. But like that is a very frustrating thing if if you're trying to get good and don't know how to get good because you don't know what axis good is on. That's a whole
big problem.
Yeah. And that and all of these make it tricky, but at the end of the day, what all of the different dev teams behind these are trying to do is to give the players the best experience.
Yeah.
And so, like for Apex, they've realized

based on how their community behaves, this is the behavior feedback we need to give.
Yes. Where for Mario Kart, because it is so incredibly casual for most of the people who play it, they can just be like, "Cool, you got some points or you lost some points at the end." You know, nobody's ranked. It's not like you have a gold rank Mario Kart. It's you see your score.
Yeah.
And you see it go up and down at the end of a game if you even care to notice the score.
Yeah.
And my guess is a majority of the people kind of look at it be like, "Oh, cool." Yeah. and
move on from that.
Yeah. And it continues to get even weirder.
Mhm.
Um and potentially less helpfully weirder, but just weirder. So, let's let's travel back towards the land of Call of Duty. We'll go into first person shooters.
Okay,
we're stick with FPS for a bit. Let's think about not Call of Duty because I don't know Call of Duty well

enough for this, but let's think about uh Team Fortress or Overwatch, right? not identical. They have more objective based gameplay than Call of Duty, which can just be Murder Ball, right? Which is fine. There's nothing wrong with Murderball, but um how how do you MMR medics? How do you MMR Mercy, right? Do they get a do they get a support MMR and then a notsupport MMR? How do tanks get MMR? Do you get MMR as a partic MMR in this case is matchmaking ranking? I don't know. or matchmaking rating. I don't know that we've defined that and we probably should. Apologies it took so long. Um, so do do you divide MMR by character? Some fighting games do that. You have a different rank as a different character because their move sets are so different and your expertise with one character may not indicate skill with another, right? But in a team-based game, the obvious answer is that the winning team gains and the losing team loses and

eventually it'll all reach equilibrium. But not everyone contributes to the winning team in the same amount or way.
And also sometimes there's fill-in positions.
Like League of Legends UKQ, you're like, "Hey, I I would like to be either mid or top,
right?"
Or, you know, I want to be jungle or support.
Yeah. And league actually at this point keeps track of your MMR on a position basis.
Yeah.
So for example, if you get autofilled, if like there are no tops and mids and to make the games go faster, you get thrown in his jungle this game. They specifically look at your rank lower [laughter] like for this. And they will actually tell people, you know, there are people getting autofilled in this game. And the amount that you not win or go up if you win, but that you lose if you lose is modified by how many people on your team get things like autofilled or others. Like their data scientists over the years have worked it into such a degree to try and make it feel not as bad

because for a lot of people playing League, they're ranking the win is the thing that they're thinking about.
Yeah. the most um during the game. But yeah, like or even like saying, you know, how do you measure supports? But also, how do you measure carries? Like, is it fair to say you're a super good carry? Because in certain games, the way that you're allowed to carry is because your support or tank or the other people on your team
are doing their job,
right? And yeah,
and it becomes so difficult because so in MOBAs, right, not all deaths are created equal.
Oh yeah.
Right. Someone on a streak is worth way more both in terms of your likelihood to win and in terms of raw gold than someone not on a streak. Right. And uh does League do the thing where your death uh bounty is based on your net worth?
Is which sorry? Does League do the thing where your death bounty is based on your

net worth?
No, it's um it's a combination of basically your your kill. You get it by doing kills and some other things. It builds it up and when you die, you lose a certain portion of your but not all of your bounties. So somebody who's been on a big streak
Yeah.
might take a couple of deaths before their bounty goes down.
Yeah. The Deadlock was experimenting with net worth bounties for a while, right? So the more that someone was above you, the more additional you got for taking them down, right? And so like it becomes very interesting. And then we look at the ways that those things um
Yeah. because like throwing themsel in front of their carry so the carry doesn't die and give the team money.
Yeah.
Is a good tactical move, right? and theoretically shouldn't penalize the support player at the end of the game with like
their kill to death ratio,
right?
But right, but also at the same time, while the support may be the person who

is most likely to contribute to a victory in that way by assisting the carry, they're the person who is least likely in some cases to be able to turn the game around if it's losing, right? And so [clears throat] everyone's impact is not the same in these games where they are fundamentally what I will say is asymmetrical. You have symmetrical win conditions but the setups for both teams are not right. You have different characters, you have different tool sets, you have different um like if one team has a fill-in jungle and the other one doesn't, that's a huge asymmetricality, right? Um, in an ideal world, of course, we'd have five teams of, uh, everyone is the same rating at their position, at the position that they're in, right? But that doesn't happen. So, we end up in this really again weird spot where it feels like
explaining all of this to a player to the degree that it feels fair that someone who played a different position gets a different rating change, especially if rating changes are public

than you did when you feel like you did better than they did and you lose or g you lose more points or gain fewer is very difficult, right? And that's why a lot of games kind of default to based on if your team wins or loses.
Yeah.
But that's also and you know, we just talked about that for a bit, but there's also like how do you rank like the freefor-all or the tier ranked finish games where you don't win most of the time? Like the Mario Karts and the Fortnites.
Yeah. you know, which is how much do you penalize or do you reward finishing the top or the top X percent or go up and down based on the quote? Like there's all sorts of ways and that's why like there isn't a single solution to these, but the general feel of these is that you want a competitive game where you don't feel like you're getting pubed all the time. And and that comes with a perceived cost to these people who are against skill-based

matchmaking. And the perceived cost of not getting pub stomped consistently is that you don't get to pub stomp people consistently.
And that's the perceived cost, right? That is that is part of it, right? That that the games feel and the term that you sometimes hear is sweaty. The games feel sweaty. The games feel very close and like you actually have to try, right? You can't just It's not a beer and pretzels game, right?
A a casual game that you could play while drinking a beer, eating a pretzel, and not really suffer in the game because of it. Um,
and that is is their desired outcome. I guess
it's funny. I think if you really wanted to make a game for that, that's in a lot of cases what single player experiences are for. you know, going and playing horror mode on Sirius Sam or, you know, something where you could just mow down
Yeah.
things and take the power fantasy with

the beer and pretzels kind of mentality because people are right. There were situations in the past where if you were a very good player, you could just like hang out in public arenas and stomp people. But we've as an industry moved away from that because it was such a bad experience for so many people.
Yeah. along the way.
And it's not beneficial when you um if especially now with Steam refunds, especially now with Steam refunds and the increasing rise of free-to-play games, right? It's of paramount importance. Someone mentioned earlier pairing people against AI uh bots for early games to get a sense of their rating. Um definitely have seen a multitude of battle royals that do that. Fortnite does that. Um, oh god, what was the 3v3 magic one that Epic was doing for a while and it shut down and I was super bummed. Um, I'm uh Apex has a whole like training mode where you can go against bots and it'll give you like pointers and stuff. Um,

and the whole point of that is that you need to get players into it enough that they're not going to bounce the first time they get their teeth kicked in. And if you don't have that system and someone pubs them in their first game, that is a there is a increased risk that you have completely lost that potential future player.
Yeah.
Like just gone.
It's interesting because like even in League of Legends, let me bring this up for a second. So there's all sorts of you can go play you can go play ranked you can go play equivalently pick up there there's multiple levels of both rank but also like non-ranked play all of which are matchmaking
there's the casual version a ram where you know it's one lane kind of random characters for everybody and you go in and it's chaotic
single favorite way to play league
yeah they still matchmake on that
but there but there is also you can just

go play versus the AI versus the computer. You can play full games versus that. And I will say, well, especially when I've been playing a lot of League, every once in a while, if I've had a couple where I lose and I just want my first win of the day, I'll go to a computer match. The single fastest matches that ever get match made for me, even faster than ARAM, are playing versus the computer.
Yep.
And that leads me to think that there are un huge amounts of people queuing just for computer games.
Yeah. specifically because they don't have to worry about the sweaty version. The computers are good enough to feel somewhat challenging. You pretty much know you're going to win. You can practice. You can play weird builds. Like nobody's complaining that, you know, you're losing top lane horribly or whatever it is because everybody's in there knowing that you're just the stakes are low.
Yeah. Although I will say uh not to devil's advocate too hard, but you do have increased parameters because um uh

an overly broad rating section won't negatively impact the game for other real players and you only are matchmaking half as many people. So,
right, it's well, it's I say it's the quickest, but that tells me at least that there are large quantities of players there. There are probably less than other cues. I don't know.
Yeah. But, you know, my guess is there, you know, if there's a couple hundred thousand people in regular league matchmaking queue at any one time that there's at least 50,000 playing just against a like,
yeah,
for for good enough numbers of people, that's a good enough experience.
Yeah.
And I've absolutely seen really good people come into bot games just so they can be like, "Yeah, I'm 22 and0, you know."
Sure.
Playing a support.
Props to you.
Yeah. and they get to have fun.
Yeah.
But they're doing it in a way, League has set that up in a way that they can do that easily without impacting other

players games.
Yeah.
But um I have something I want to tangent to that is related to this will.
Yeah.
Before I do, is there something else you want to say on this thread on this line? Well, so I mean we have two big philosophical questions and I like tackling yours first and then I mean I guess mine sort of relates to that which and my big question is um if winning is the only fun part of the game like if winning matters so much.
Oh, you're gonna take mine? Go for it.
Oh yeah. Well, I was gonna I was going to do this and then let you lead into the conclusion to this, which is um is the game actually fun,
right? Like in the middle when Will and I were talking in our pre-show about this, I just kind of came up with a rant that I was that I hadn't really thought about much, but this made me think about which yeah, you know, if winning is the fun part of the game or you know, is the

winning part fun or excuse me, is winning the fun part or is playing the game the fun part? And if winning's the most enjoyable part of the game, is it actually a bad game? Because um ideally the game's enjoyable enough that winning or losing is secondary to the gameplay itself. And let me give you a good one of my favorite examples of this is even though I haven't played Halo in at least a decade, multiplayer [snorts] Halo for a while where you know it's like capture the flag and people are running around getting the flag and taking it back and forth, but there's also people skirmishing on the side or hunting, you know, like, oh, I found the tank and I unlocked it or here's where the rocket launcher spawns and they're just everybody's running around doing looser stuff that still contribute to the game, but I'm going going to die 12 times in that and quickly respawn and come back in or like in Dest Destiny 1 multiplayer with all you know the big team battles back and forth and just

kind of going through that it felt like they'd focused more on the game play and that the actual like there wasn't the big ranked competitive system or other stuff. it was let's get in and just you know have a good slaughterfest and a like they'd really emphasized that part of the game to the point where death was usually I won't say trivial but you just knew you were going to die a lot during the game
and it wasn't single elimination counter strike or you know the economy behind league or some of these other things and not saying that that those are bad things but it feels like a lot of games have shifted the focus so much on to the W that they're not paying or at least the community has that the big part of the game is all in support of getting the W as opposed to just having a good time around that.
And I do wonder if part of that is that

the ceiling for skill in all of these competitive multiplayer games of all of these different types is now higher than it ever has been. So the floor remains the floor. You can have never played one of these games before and be bad, right? I myself am bad at many games. I am bad professionally at games, right? Like there's nothing wrong with being bad at games. Cool, great, awesome. But some people have played that genre of game since it came out, right? Some people have been playing a firstperson shooter like Halo every day since Halo 1 match m Halo 1 multiplayer came out and they are really flipping good at Halo multiplayer and they are better than it was possible for the best player to be or the the best player was back in the days when Jeff was playing. And I do wonder if that discrepancy in skill and the increased level of the top end is part

of what drives the W or rather that the because the ceiling is so high and people want to climb towards the upper echelons, right? They feel as though and this is interesting. They feel as though witting is the acquisition of skill rather than the test of it. And I think that might be part of it, right? So, and by that I mean
it's a fun line.
Yeah. People believe that they are better because they won that game of League of Legends. And that is
backwards. They won that game of League of Legends because they have gotten better. And so the there's this topsyturviness to it which I think is part of the reason that the the the win is so important now. But I also feel like a lot of games have lost some of the momentto- moment fun feel

because they're so focused on the winning and losing nature of the competition.
Yeah.
And that there are larger game like I feel like Fortnite has managed for example to keep it in kind of the weird zany way. And it's because also you're not expecting to win for most players, right,
terribly of like you don't get the chicken dinner terribly often.
You're in there with x number of other people, you know, doing as well as you can, but also doing weird
That is part of the game.
Yeah. And then and that's and that's the point, right? You're there to see who would win in a Fortnite match, Goku or Superman. Like this is fine, right? That's that's the buyin, right? Um, I did have another and and and this applies to all kinds of things, right? One of the and I think it applies outside digital gaming as well. So, in in Magic Arena below

bronze, there are spark ranks. And the spark ranks are your placements. They're where everybody gets put when you first install Magic Arena. You are brand new. You're you get here's your spark rank, right? Be in spark rank. Okay, great. I saw a screenshot of someone with a spark rank playing and I don't know whether it was a smurf or just a new player who was net decking who was playing as one of their first matches in the game, not the pre-made deck that the game gives you for free when you start and is designed to help you understand the fundamentals of the game. They were playing like a full net decked tier one deck that they could not have had resources to build then. So, must have spent money as one of their first things in the game to build a tier one deck and then play it in the unranked placement matches at the very bottom below the bottom end of the ladder.
What? Like,

they're coming into the digital game with that, right? Whether it's
they're bringing a Mark Robear six foot long Super Soaker to a match where everybody's using little water pistols.
Correct. And I think that that is so I my my my thought here is only that I think that exhibits a feeling of competition out that this is not a purely digital games problem. I think this is also this is a games as a whole problem. We just have a very specific lens that we can look at it through and that is that digital games uh experience this phenomena and that's one of the reasons that I think skillbased matchmaking is so crucial.
Yeah.
Yeah. Yeah.
And it's also become so ingrained
at this point and I understand the feeling that like people are asking for for the kind of that nostalgia factor of, you know, remember when it was more casual and
so and so, but
eliminating skill-based matchmaking isn't what's going to get us there.
Agreed.
I think it's designing games that have

more of the, you know, at the end of one of those old Halo games, I didn't really care if we won or lost because we had such a good time playing. we'd finish it up and immediately queue for the next one to get back into the chaos.
Yeah.
And so it's building the games with that looser sense of the victory being the thing, but it's more of getting it's that last it's the countdown of points and watching, you know, your teams go head-to-head getting them and all of the stuff catching up and the craziness around it.
Yeah. I I wonder if um one of the things that uh contributes to that um is uh the chaos element, right? Halo wasn't incredibly chaotic, but it was a bit chaotic, right? And and by chaotic, I don't mean there are bullets whizzing everywhere. I mean matches never turned out quite the same. Blood Gulch didn't play the same way every time from what I remember, right? Because people tried different tactics and preferred different weaponry and so on and so forth. just the mix of different people who all wanted different things.

Yep.
Would cause it to
Yes. And um I feel like the chaos inherent in in some games either by having like a third team, right? So like a three team thing like what Warogs is doing. Warogs is currently taking over Twitch. It's a um milim three team shooter where pretty much everyone is playing a role. There are very few if any like support NPCs. Even the people who are flying helicopters to bring people to the front lines are generally players. Kind of cool. Um
more like the old planet side.
Yeah. It's three teams, right? Which is inherently chaotic because if one team gets too far ahead, then the other two teams will band together and beat them up. And that applies to like Final Fantasy 14's front line, which is a PvP murder ball nightmare that I love fiercely with every fiber of my being. And not just because I have hundreds and

hundreds and hundreds of matches of Front Line under my belt. Um, I think the I think there is the the whimsy returns when we bring back weird, right? Like
Yeah. And I feel like
weird and asymmetric or purposeful imbalance like I mentioned Planet Side. Planet Side was also three player or three teams
and actually was designed by a guy who designed a game called Tannerus before that which was 4v4 or four teams of four tanks all battling. And on both of those games as you you took control of territory, but that extended your team and meant that for the other two or three teams, it was easier to push in and you had more to defend. So
the states bounced back and forth.
Yeah. And it it lent a certain air of chaos to it in a good way because you your team doing well set you up to be targets and return you to a more of an equilibrium state

but still got you points towards like the victory.
Yeah. [clears throat]
Yeah. And I think that's actually one of the reasons why Friendlop has been such an has taken off so much is those games are purposefully designed to be more sloppy and more chaotic in their play style. And I don't think I've seen almost any of them that are like competitive head-to-head. But it it brings back some of that lobby feel of getting matched in with wacky people or wacky the wackiness ensuing from the old style of servers with a little bit of the jank and some of the play is that for a lot of friends love games winning or losing the game is not nearly as important as the experience along the way.
Yeah.
The fun is in the game not necessarily in the victory or the loss.
Right. I think there are there are some that have like defined win conditions like peak, right? Peak is get to

you'll never guess the peak. Wow, amazing. Um and that that's the victory condition, right? But like if everyone dies because Timmy threw the backpack off the edge, that's almost a better story than everything went perfectly and then we won, right? Like
losing the backpack is indeed peak gameplay,
right? Yeah. Yeah. Exactly. In every sense of the word, right? And I think that that I I agree. I think that that style of like fun, ridiculous multiplayer that provides worthy adversaries in the form of the computer without demanding that other people suffer for your nonsense is is um
uh you know, one of the reasons that it's so popular. Yeah. Like almost watchable said it's the win condition of friends slop isn't make everyone else lose which helps a lot. Yeah.
Yeah. And it was just like that is I

think I feel like friends slop is a specific response in a weird way to online multiplayer getting as refined as it is.
Yeah. is it's it's bringing back some of what older gamers used to feel and newer gamers never got to experience because for all of the good that the multiplayer matchmaking does like of making games feel like you actually have a chance and you're not losing all the time from a set of trolls sitting in and smurfing. They also do take out some of the chaos factor.
Yeah.
That came with that.
Right. and and maybe it's just that COD needs a better casual mode
like maybe
or a way to refine the matchmaking to fulfill that purpose.
That is the hard part, right? That is that like it's not that there shouldn't be skill-based matchmaking. It's that we're consistently looking for better ways to do skill-based matchmaking. Let me give you an example. Remember when I

was talking about arena and I said that there were two factors? I goofed. There are actually three
and it depends on the game mode you're playing. Um, in straight constructed um, one v one, there isn't one. There only two, your MMR and your rank. In limited, which is draft, it's have you, how many wins and losses do you have right now, right, in your draft run. So, if I'm three and two, I'm likely to get placed with someone who's three and two. And if you're playing brawl, which is their commander equivalent, not competitive brawl, regular brawl, competitive brawl is 1v one constructed. But if you're playing casual brawl, it will try and match you with a commander of the roughly similar power level because even magic with 1v one has recognized that there needs to be a place where you can just blow off steam and do random stuff. Now, it's not perfect. Not saying it's perfect. I'm saying the idea is good, right? That there is a way to use matchmaking to accomplish something like

this result, right? Or maybe you can tick a box, right? when you like go in and say what I'm looking for is this broader experience, right? And the matchmaking can add additional factors like your commander and brawl that will help with that. But again, I think the solution is better and more specialized matchmaking, not let's just throw skill-based matchmaking out the window and make casual play casual cues an absolute cesspool. Because my read is that especially in a game that is as has such a storied history of colorful language and interesting characters as Call of Duty does. I feel like removing skill-based matchmaking dooms casual playlist to be an absolute cesspool.
And it's funny, I was thinking about it a little bit. It's like you could for example make make one of the matchmaking cues say have a title of something like you know unranked or non-match or you

know non-skllbased. We'd come up with a sexy term for it.
Yeah. But I have a feeling that 90% of the people who click on that would all be a certain level or above
and they'd be like self- selecting into the same kind of thing because everybody else would try it once be like nope
this is not a good experience and they'd go back to the regular queue.
Yeah. Like I feel like experiments on that would show people that the there's a large amount of people who don't want to go in and basically be the cannon fodder for the certain amount of other people.
Yeah, it's kind of like the self- selecting um cheaters queue,
which some games have. Some games have a instead of banning you when they find out you're cheating or manipulating your rank or doing any number of unsavory things, they just put you in a special queue with everyone else just like you. Have fun. H
enjoy that.
It's my second favorite way that people

deal with cheaters. My favorite way is the way Deadlock does it because if someone's detected cheating in Deadlock before they get permabanned, they get turned into a frog and everyone on both team gets to attack them and then they get banned which is great. [snorts]
Wow.
Yeah, it's so funny cuz like you literally did this to yourself. Like I don't Okay, great. Um yeah. Uh wondering if you could make a matchmaking system that was designed to make one team more or less guaranteed to win every game and then change rating based on how long the game lasted.
Depends on the game. You could.
Yes, I think you could. I think it is far more likely that that would occur in an asymmetrical game. And I do I know a fair bit about Dead by Daylight because I've watched a lot of Dead by Daylight, but I haven't played it. Something tells me that a system like that would be possible. And that it's probably the

most well-known asymmetrical game right now in that one team is one person and the other team is four people. They have different victory conditions. They have different actions they can take. It's completely asymmetrical, right? Um they even have survivors are in third person and killer is in first person, right? Like
it's very asymmetrical. Um and that is kind of how that works, I would I would say. Um, but I think you could do it with other games, too. I I think that is a matchmaking challenge that could be risen to. It would just require a lot of data science and you would need to tell people you were doing it. I think that's really important
because it well I could absolutely also see a mode and I I don't quite know how we'd pull it off because this is 10 seconds of thinking about it but do you remember the old sidekicking system will in um City of Heroes where
you could basic sidekicking and mentoring where you could go in and it would basically either raise or lower your level to match with the party so

that your 30th level guy could go play with the new group of 10th level people and it would lower your level enough.
Yeah. And not necessarily saying that on matchmaking, but a way to basically purposefully go in and play with better players or players who weren't as good as you so that they could kind of watch what a better player does.
Mhm. in some way like because I actually had this in League of Legends when I started playing some of the people I played with had been playing for a while and getting to like lane with them or listen to them on voice chat talk about oh we should be rotating here, you know, quit hiding in the bush as a support. You're actually supposed to be doing something, you know, besides just healing like help me realize things faster than I otherwise would have. Yep.
And upped my gameplay faster. So, I was lucky that I got to play with more experienced people and that helped. And it's it'd be fun to find a way to replicate that. But also, I don't I

don't have an immediate solution.
Hott has one. Hott has a character that's literally two players playing the same character, Cho'GL. One person plays Cho, one person plays Gaul. One person's the legs and one person's the um offensive power essentially. Uh, and if if you're ever playing uh Heroes of the Storm with a new person, having them play uh the part that's doing the damage in Shogal is like choice cuz they don't make any of the decisions, but they're bound by the decisions about like where you go and when you rotate and all of that stuff. But again, that's that's cuz Heroes of the Storm is really weird. Capital R, capital W. I'm very glad it's making a comeback because we're definitely gonna need to have a design club about how weird that game is. Um
Yeah.
But outside of something like that, I think it's really hard to
Yeah. But again, some of those are also goals you can have that like as your when your game hits a certain level, like cool, we could start thinking about de devoting resources to things like

this.
Yeah. Yeah. Agree. But yeah, so I think a lot of the summary is it's an understandable thing to say. I want a more casual play or I want to feel like if I'm a really good player, I should win a lot more often, but the cost for that is that for a majority of people, the games get worse.
Yeah. And also you may think you're really good, but also s if those things get turned off, you also might find that you are one of the people getting pub stomped,
right?
And that um that's that would in some cases be doubly deflating.
Yeah.
Because you're going to be coming in being like, I'm so much better
and then you're not
and then still proceed to get stomped. Like
you find out how big the pond really is and how how small a fish you really are.
Yeah.
Yeah. And that just that makes it really tricky to make these games what the large goal of which is

fun for lots of people.
Yeah. And I think I think it's really important to acknowledge too in all of this matchmaking is not perfect and will never be perfect but it is better than um it is better than not doing it and it's better than most other solutions that we've come up with so far I would say. Um, and it's definitely definitely definitely better than free-for-all. So,
Mhm.
I think that is a good place for us to wrap our discussion. Uh, we were talking a little bit earlier about Friendlop, a recently emerging genre. We're actually going to be doing a design club next week on another recently emerging genre of game, tidyup. going to chat with actually a group who's who we've had on before, but that is currently making a tidyup game that's already been even though it has just been listed on Steam.
Mhm.
Um has already won a couple of awards or been put in categories for this is

really good and I think we're getting them on launch week.
I think that sounds right
or right around that. But for tidy a tidyup game which is a looks really cool and b is a category that literally just formed like I think the library was the first acknowledged like one. So we're in the stage where people have seen that and said oh I could go make we're hitting that first wave of games of people saying I've seen what you've done. Let me give you know let me take it and expand my
thoughts or my studios feels on it.
Yeah absolutely. I'm I'm super looking forward to it. I think it's going to be great. Uh uh brief scheduling note as well. There is no breakfast this week. Matt is away. Uh there will not be breakfast on Friday next week either, but there may be a breakfast uh sometime. Matt and I have to schedule that. So, please keep your eye out on the upcoming streams tab. We'll let you all know when breakfast will be. Uh I think that's everything, which means that the only thing that is left is say

bye to the people, Jeff.
Bye to the people, Jeff. Thanks everybody. Have a great week. [music]