Your thoughts on Engagement Based Matchmaking?

Alek · Feb 8, 2020

Whether or not a game should feature skill based matchmaking is a pretty common debate right now. However, something that I think is affecting the way the industry is moving, is called engagement based matchmaking (EBMM). I wanted to use this thread as a platform to talk about it.

What is engagement based matchmaking?

Engagement based matchmaking is a matchmaking system that places players together based on whether or not the matchup is likely to cause players to churn, rather than whether they're the same skill.

For the most part, engagement based matchmaking works in a similar way to skill based matchmaking, but it deviates when it identifies a scenario that is likely to predict the player quitting the game. The typical scenario is if you lose several times against players at the same skill level as you, the game throws you a mismatched game so that you are almost guaranteed to perform well.

The goal of EBMM is to encourage engagement, to keep people playing the game, both for longer play sessions, and more frequently across a long period of time. The most important attribute is to prevent players from churning, to help ensure that they come back day after day.

Where can I read about EBMM?

There are a couple of papers that outline how EBMM can be applied, though these are fairly technical. You can also just search churn prediction in scholar, as many of those articles raise that losing several times predicts churn.

https://arxiv.org/pdf/1702.06820.pdf

Read this first one if no others.

Early churn prediction with personalized targeting in mobile social games

Customer churn is a widely known term in many industries, including banking, telecommunications and gaming. By definition, churn represents the act of…

www.sciencedirect.com

Churn Prediction in Mobile Social Games: Towards a Complete Assessment Using Survival Ensembles

Reducing user attrition, i.e. churn, is a broad challenge faced by several industries. In mobile social games, decreasing churn is decisive to increase player retention and rise revenues. Churn prediction models allow to understand player loyalty and to anticipate when they will stop playing a...

ieeexplore.ieee.org

US10286327B2 - Multiplayer video game matchmaking system and methods - Google Patents

Embodiments of systems presented herein may identify users to play a multiplayer video game together using a mapping system and machine learning algorithms to create sets of matchmaking plans for the multiplayer video game that increases player or user retention. Embodiments of systems presented...

patents.google.com

What's perhaps interesting to note is that the data from the first article here indicates that skill based matchmaking actually does not reduce churn. In their study, players were actually better retained with randomised matchmaking than with SBMM. While at face value SBMM often seems better for the community on paper, that might not be the case, at the very least, the data doesn't point to a clearcut difference between the two.

Instead, I suspect the reason that game developers are pushing towards what they label 'skill based matchmaking' is because of EBMM. Crucially, EBMM is dependent on skill based matchmaking. In order to adjust the difficulty of the experience in a multiplayer setting, the system must know the skill of all players. The significant distinction is that as and when appropriate, the system will manipulate the gameplay conditions to avoid player churn.

My thoughts...

Speaking anecdotally, the conversations I have with both developers and academics point towards EBMM as the future for how players match together in online multiplayer environments. In fact, the conversations point to this type of system already being at play in most of the bigger modern competitive games, as of around 2017.

Ultimately, the developers of these hefty large service games don't see skill balancing as the priority. Instead, they're interested in keeping people playing, and keeping people spending for as long as possible. EBMM helps the studio do that better than SBMM or connection based matchmaking.

As to whether this is ethical, I wanted to make this topic to get people talking about it, so I won't frame the discussion by adding my views on that to the OP. Nevertheless, it would be good to hear everyone's thoughts.

bananabread · Feb 8, 2020

I've always been suspicious of this being a thing based on my own personal experiences. Far too many DotA/League runs of 'win 3-4 games in a row, get matched with raging lunatic feeders for the next 3-4 games, repeat forever without ever progressing' over the years. It's easy to start to believe you're cursed or that the matchmaking system is deliberately working against you.

Juan · Feb 8, 2020

Will develop later as on my phone, but I do think it's an interesting concept as you can easily get players frustrated by losing games even if it was a close call due to matching with similar skilled player, this way, it helps keeping a good mood feeling, even if artificial.

As to whatever it is ethical or not, I would say it depends on your transparency about it. Acknowledging players had a hard time and telling them the game is adjusting its matchmaking to find more enjoyable games allow the players to understand what's happening and still being honest about the game trying to match the players for a more enjoyable game.

I still think random matchmaking can always perform better that said.

Hella · Feb 8, 2020

EBMM just sounds like a bad, manipulative system designed to keep people hooked rather than give them the best matches possible. If it takes off, the next step would be to microtarget players that spend money and improve their experience (y'know, give them teh wins) at the expensive of non-paying players; so they keep paying, y'know.

It's an incredibly toxic way to curate a community, but hey if it makes a buck I fully expect everyone to embrace it.

collige · Feb 8, 2020

Hella said:
EBMM just sounds like a bad, manipulative system designed to keep people hooked rather than give them the best matches possible. If it takes off, the next step would be to microtarget players that spend money and improve their experience (y'know, give them teh wins) at the expensive of non-paying players; so they keep paying, y'know.

It's an incredibly toxic way to curate a community, but hey if it makes a buck I fully expect everyone to embrace it.

Yeah, this pretty much. I don't want to be condescended to by a game and if they actually publicized when it was working I suspect most players would feel the same.

"To our best knowledge, we have not seen any existing matchmaking method that formally treats matchmaking as an optimization problem to maximize player engagement". Yeah no shit because that flies in the face of playing a skill based game in the first place.

Kittens McMittens · Feb 8, 2020

So, let me get this straight. If I rage quit or stop playing immediately after a loss, eventually I'll just start getting rotated through lobbies of lower skill players for me to stomp on? I just have to train the matchmaker to know that I'm a massive baby who just wants easy pubstomps?

Einchy · Feb 8, 2020

I hate all this type of shit, just find me a lobby with the best ping.

Clay · Feb 9, 2020

Would definitely get me to play fewer multiplayer. Manipulative shit like this is a big part of why I quit social media. I'm not going to spend money on a game that uses research-based tactics to keep me mindlessly playing as long as possible.

Alek · Feb 9, 2020

Clay said:
Would definitely get me to play fewer multiplayer. Manipulative shit like this is a big part of why I quit social media. I'm not going to spend money on a game that uses research-based tactics to keep me mindlessly playing as long as possible.

'User research based tactics' is the wrong term here. User researchers endeavour to provide the voice of the player experience, user researchers are the ones communicating the players frustrations back to the development team.

This type of design choice is driven by data science / marketing.

Juan said:
Will develop later as on my phone, but I do think it's an interesting concept as you can easily get players frustrated by losing games even if it was a close call due to matching with similar skilled player, this way, it helps keeping a good mood feeling, even if artificial.

As to whatever it is ethical or not, I would say it depends on your transparency about it. Acknowledging players had a hard time and telling them the game is adjusting its matchmaking to find more enjoyable games allow the players to understand what's happening and still being honest about the game trying to match the players for a more enjoyable game.

I still think random matchmaking can always perform better that said.

The problem is, that if you are transparent about this kind of system then I suspect you see a diminished impact of EBMM. It relies on players thinking they're suddenly doing better at the game, not acknowledging that the system has thrown them a free one.

Clay said:
Would definitely get me to play fewer multiplayer. Manipulative shit like this is a big part of why I quit social media. I'm not going to spend money on a game that uses research-based tactics to keep me mindlessly playing as long as possible.

I think this system is already at play in the majority of major competitive games. So unless you're already feeling the impact then you probably won't feel like you're being affected by it.

Sanka · Feb 9, 2020

Nah, sounds horrible but it's exactly what companies would want to achieve so I am not suprised that they are hyped about it. Just skill based matchmaking is good enough.

But this also only seems to apply to casual not ranked games, right? In that case I couldn't care less. Actually, who cares about matchmaking in casual games?

Clay · Feb 9, 2020

Chronospherics said:
'User research based tactics' is the wrong term here. User researchers endeavour to provide the voice of the player experience, user researchers are the ones communicating the players frustrations back to the development team.

This type of design choice is driven by data science / marketing.

I said "uses," not "user." I see data science and analysis as a form of research, and marketing departments obviously conduct a lot of research. Not really sure what you're point is here, seems like a bizarre attempt to unnecessarily argue semantics.

Chronospherics said:
I think this system is already at play in the majority of major competitive games. So unless you're already feeling the impact then you probably won't feel like you're being affected by it.

What games do you think use it? Rocket League is the only competitive game I've really gotten into in the last few years, micro transactions and tedious mechanics like the equipment and character leveling in many shooters has already turned me off from most multiplayer games, this would just be another reason to avoid them.

Alek · Feb 9, 2020

Clay said:
I said "uses," not "user." I see data science and analysis as a form of research, and marketing departments obviously conduct a lot of research. Not really sure what you're point is here, seems like a bizarre attempt to unnecessarily argue semantics.

No, sorry I just misread your post (as you suggest I read user instead of uses) I'm used to seeing 'user research' get blamed whenever there's poor game design that's linked to manipulative psychology, and as user research is my job title I tend to try and explain what we do, and don't do.

I'm not against the idea of leveraging data analytics to understand what your players are doing though. I don't think churn prediction is inherently a bad thing. For instance, you might see that experiencing a certain bug is predicting churn and that might prompt the dev team to prioritise that issue, or you might see that players are quitting more often after playing a certain game type which might lead you to look into why that's happening and improving areas of the design of said gametype.

I think the problem EBMM isn't that it's research (or analytics driven) but that it seeks to manipulate the player. It's a cheap trick to keep you playing, and I think it also promotes potentially harmful patterns of play.

It's also the developer taking the easy way out with the games design. Instead of focusing on means in which the game could be more fun and engaging even when you're having a bad match (e.g. give me better feedback when I die, making the death state part of the experience in some way, offering modes where players can relax and play without stress).

I don't know if excessive video game play can be classified as addiction (would rather not debate that) but I think that it excessive video gameplay can definitely be a problematic, harmful behaviour. I have seen it first hand, and from players in many instances. I think this type of manipulative design encourages that.

Clay said:
What games do you think use it? Rocket League is the only competitive game I've really gotten into in the last few years, micro transactions and tedious mechanics like the equipment and character leveling in many shooters has already turned me off from most multiplayer games, this would just be another reason to avoid them.

Off the top of my head I think something similar to this is at play in Apex Legends and Fortnite, as well as many mobile titles. I think in Fortnite and Apex Legends, this is what the data from the post match feedback also contributes to. Apex asked you if you had a good time in your recent match, Fortnite has some thumbs up system, I suspect that data is being used so that the developer can understand churn, and I think the recent introduction of skill based matchmaking into these games, is so that they can reduce player churn by getting ahead of instances where churn is predicted (i.e. if 5 straight loses predict churn, throw a guy into a low MMR lobby when he hits 3 or 4 until he wins.

I think it's also likely at play in titles like Overwatch, Rainbow Six Siege and Rocket League. But who knows! These games all have 'SBMM' applied to their casual matchmaking, and there's not a strong reason to do that unless you're looking to use those player performance metrics to get ahead of churn. So I think it's likely something similar is at play in these games too.

It's not just about messing with the matchmaking system though. It's likely also loot distributions and other mechanics. For instance you might have data that predicts players don't buy more loot boxes after they open 10 without getting a box, so you make it so a rare item is guaranteed after 9 box openings. Sometimes (like in hearthstone) these systems are transparent, and I think that's fine (because it lets the player factor that 'guaranteed chance' into their decision making when making purchases, but in most cases this data isn't presented to the player.

I think it brings up a whole heap of ethical issues, and it's an area that needs some kind of self-regulation. It comes down to transparency. You should be able to make fun, engaging experiences while being honest to the player, and that's a failing of the design that needs addressing through... better design, but instead what we're seeing is system after system that seeks to warp the players perception of their experience.

The developer wants you to think that you're doing better at the game than you are, that you're luckier when you buy packs than you truly are. There are honest approaches to improving the players sense of competence, and reward, and then there are dishonest ones, and what were seeing with EBMM is the latter.

Pellaidh · Feb 9, 2020

I hate it.

The most obvious example of something like this ruining a game would be how Fortnite handles bots. Which isn't quite the same, but is still relevant to the discussion. Basically, if your MMR (or whatever the game uses to gauge your skill) is low, the game will match you with a random number of bots. These bots are absolutely terrible, and exist only to give you free kills. This is supposed to help new players learn the game, but these bots are so bad they don't really do that. Instead, they actually make winning feel super unrewarding, because you never know if you did well because you're good at the game, or just because your lobby had a bunch of bots.

EBMM has basically the same problem: you can't know if you're winning because of your skill, or because the matchmaking just placed you in a game you're supposed to win. Except unlike the bot example in Fortnite, this is actually way worse, because it's matching you against other people. Meaning that for every game you're supposed to win, somebody else has been put into a super unfun match that they have no chance of winning. And that same situation will eventually happen to you too, meaning that some of your games will inevitably just be gigantic unwinnable wastes of time. As someone with not a lot of time to play games, having some of that time be wasted on getting stomped simply because the game wants me to get stomped is not exactly my idea of fun.

For me, the only really fair way to do matchmaking is to be fully transparent with it: both by documentating how the ranking system actually works, and by showing the player his MMR score. Sadly, the games that actually do this are in the vast minority.

I'm also wondering how much simply adjusting the MMR algorithm would help with this. If your end goal is to stop players from losing multiple matches in a row, then simply adjusting the MMR calculation so that ranking decays more quickly with consecutive losses should help with that. It's not an ideal solution because you probably don't want your ratings to be super volatile, but has there been any research done into that? Because the linked papers don't really seem to touch on this at all.

Alek · Feb 9, 2020

Pellaidh said:
I hate it.

The most obvious example of something like this ruining a game would be how Fortnite handles bots. Which isn't quite the same, but is still relevant to the discussion. Basically, if your MMR (or whatever the game uses to gauge your skill) is low, the game will match you with a random number of bots. These bots are absolutely terrible, and exist only to give you free kills. This is supposed to help new players learn the game, but these bots are so bad they don't really do that. Instead, they actually make winning feel super unrewarding, because you never know if you did well because you're good at the game, or just because your lobby had a bunch of bots.

EBMM has basically the same problem: you can't know if you're winning because of your skill, or because the matchmaking just placed you in a game you're supposed to win. Except unlike the bot example in Fortnite, this is actually way worse, because it's matching you against other people. Meaning that for every game you're supposed to win, somebody else has been put into a super unfun match that they have no chance of winning. And that same situation will eventually happen to you too, meaning that some of your games will inevitably just be gigantic unwinnable wastes of time. As someone with not a lot of time to play games, having some of that time be wasted on getting stomped simply because the game wants me to get stomped is not exactly my idea of fun.

For me, the only really fair way to do matchmaking is to be fully transparent with it: both by documentating how the ranking system actually works, and by showing the player his MMR score. Sadly, the games that actually do this are in the vast minority.

I'm also wondering how much simply adjusting the MMR algorithm would help with this. If your end goal is to stop players from losing multiple matches in a row, then simply adjusting the MMR calculation so that ranking decays more quickly with consecutive losses should help with that. It's not an ideal solution because you probably don't want your ratings to be super volatile, but has there been any research done into that? Because the linked papers don't really seem to touch on this at all.

Usually I think this type of thing is applied to casual games so MMR is completely intransparent. With that in mind, MMR could be super volitile, yes. I think that would create a very similar effect, as you suggest. Lose a few and you get super easy games, win a few and you get super hard games, until you start losing again. You'll spin back and forth and likely never really learn anything because the strategies that work in one game will not work in the next.

I think if it were super volatile in any kind of ranked style mode, players would get incredibly frustrated seeing their ranks depreciate so quickly. I think it's actually possible to identify a players skill very quickly, but ranked modes nowadays exist for the sake of the grind, not as a means to identify a players skill level. It's all about getting people invested for long periods of time.

Ruken · Feb 10, 2020

I'd rather grind out good games in a flat ELO system, knowing my time investment is fair and that the thing keeping me interested is the feeling of improvement and the contribution I'm having in the game, rather than chasing some "high" like the system in the OP.

I read something else last week on the overwatch university subreddit that alluded to this.... That that's why people (most likely subconsciously) chase kills instead of chasing wins. Getting kills is the easiest way to have fun with the game, rather than doing something that is conducive to winning the team fight or game as a whole. Granted sometimes those two things line up by happenstance. You just don't have as big of a personal impact if it's engagement based matchmaking, so you have to make your own fun.

The most fun I ever had laddering was in the old league of legends system with visible MMR.
You tended to stick more around the MMR you were performing at, until you legit improved or you had some loss streak because you were in the wrong headspace.

It didn't have the big swings, so when I was ready to buckle down and be serious, I didn't have 200 points to grind through, and could reliably go upward
I could do my bullshit teemo stuff and stick around 1500-1550, then when I was falling too far or ready to make a serious push to move up, I could do it by buckling down and playing the restricted champion pool that I was best at.

TheUnseenTheUnheard · Feb 10, 2020

Okay so how do we stop this from being implemented?

Alek · Feb 10, 2020

TheUnseenTheUnheard said:
Okay so how do we stop this from being implemented?

I don't know. Complaining about this type of thing feels like this, because no one cares.

Everyone seems to consume everything at such a superficial level that it's difficult to complain about detail. Like how do you stop loot boxes being rigged either? It's very difficult. It's easy for a developer to pull strings behind the scenes and because there's no regulation on these issues, there's very little anyone can do.

Players just need to push for transparency between them and the developer. I think we should get to know how the match making is works. We should get to know when we're playing against bots or human players, and monetised elements that revolve around chance (e.g. loot boxes, card packs etc) should display their odds.

Sly Chimera · Feb 10, 2020

Hella said:
EBMM just sounds like a bad, manipulative system designed to keep people hooked rather than give them the best matches possible. If it takes off, the next step would be to microtarget players that spend money and improve their experience (y'know, give them teh wins) at the expensive of non-paying players; so they keep paying, y'know.

It's an incredibly toxic way to curate a community, but hey if it makes a buck I fully expect everyone to embrace it.

Oh that already exists too haha. The patent was like buy something with micro transactions then get to play horrible people to make your purchase worth it

Juan · Feb 10, 2020

Chronospherics said:
The problem is, that if you are transparent about this kind of system then I suspect you see a diminished impact of EBMM. It relies on players thinking they're suddenly doing better at the game, not acknowledging that the system has thrown them a free one.

Well, that's an assumption. This would need to be tested in a close environment to see if it indeed impacts the way player are engaged about it and think about the game they are about to play next.

As I think you said, Apex Legends is, I suspect, doing something similar, since they often ask how I've been enjoying my latest game. Usually, when I say I didn't, the match after is quite better in term of score I do, and I think this is transparent enough. Even if Respawn doesn't tell "The next game is going to be better for you", they still aknowledge my answer influenced the matchmaking, and it doesn't seem to negatively impact my engagement with the game, quite the opposite in fact.

Of course, I'm one of many, so this would need to be tested once again, but I definitely think you can have transparency and positive impact regarding EBMM.

Alek · Feb 10, 2020

Juan said:
Well, that's an assumption. This would need to be tested in a close environment to see if it indeed impacts the way player are engaged about it and think about the game they are about to play next.

As I think you said, Apex Legends is, I suspect, doing something similar, since they often ask how I've been enjoying my latest game. Usually, when I say I didn't, the match after is quite better in term of score I do, and I think this is transparent enough. Even if Respawn doesn't tell "The next game is going to be better for you", they still aknowledge my answer influenced the matchmaking, and it doesn't seem to negatively impact my engagement with the game, quite the opposite in fact.

Of course, I'm one of many, so this would need to be tested once again, but I definitely think you can have transparency and positive impact regarding EBMM.

I don't think the data from those surveys is used that way. I think they use that data to predict which conditions lead to a bad experience. It's likely that when you click no, it just overlaps with when you had a bad time. I don't think asking you if you had fun in the previous game is a transparent way of throwing you a free game.

For what it's worth, I haven't been given a post-match survey in the past 200 or so games. I don't think that means my matches aren't affected by some sort of EBMM. It's really hard to understand what they're doing with our data or the matchmaking without any transparency, we can only guess.

My best guess though, is that you're seeing a lot of live service games introduce EBMM right now, and the reason for that is that there's no financial incentive not to. It reduces churn, and players aren't aware that it's happening. I think if players were aware, they might have a different experience, but as you say it's hard to know. I don't think that the presence of a post-match survey provides transparency on this though, far from it, it just leads to more questions on what that data is doing.

Deleted member 17643 · Feb 10, 2020

As others have said, I suspect Apex uses this in some form. We'll go from lobbies full of diamond/apex level players to those who might as well be AFK.

Juan · Feb 10, 2020

Chronospherics said:
I don't think that the presence of a post-match survey provides transparency on this though, far from it, it just leads to more questions on what that data is doing.

I don't think average players think more about this survey than just being a simple survey. You're a User Research and I'm a Product Designer, so we're definitely biased regarding how data are used in such context, but, and it's totally a personal opinion, I don't believ average players think that far about it.

Alek · Feb 10, 2020

Juan said:
I don't think average players think more about this survey than just being a simple survey. You're a User Research and I'm a Product Designer, so we're definitely biased regarding how data are used in such context, but, and it's totally a personal opinion, I don't believ average players think that far about it.

I don't disagree with that at all. But I've yet to find anyone that if they were made aware, would be happy if they were told they were being thrown an easy game. I think it diminishes the sense of reward, which defeats the point.

This type of model coincides with self-determination theories of player motivation (which are also very popular within the industry), one attribute of this theory is idea that developing a sense of competence is integral to sustained engagement. Experiencing several losses in a row challenges that sense of competence and sees players lose motivation to play the game. That's where EBMM comes in, restoring that sense of competence by throwing the player an easy win, without telling them they've stacked the deck in the players favour.

I might be wrong, but I suspect transparent EBMM would diminish that restoration of competence. and in turn end up being significantly less successful at reducing churn. I think that this system relies on the fact that it pulls the strings without the players awareness.

Though, it would be good to see a study comparing randomMM, SBMM, EBMM and transparent EBMM.

psilocybe · Feb 10, 2020

Interesting. It is the first time I read about EBMM. It is actually what I experience with Modern Warfare. I play against pretty good people, get wrecked, then I'm placed in a visible easy lobby.

It sucks because I can see the pity the matchmaking is having on me. I would prefer a decent SBMM where matches are intense and balanced.

Filipus · Feb 11, 2020

I will always prefer "free lobbies" and then some form of ranked play.
Basically prioritize ping, then skill (just have a hidden elo or something).
Then allow for users to play ranked with a "real elo".

I also believe matchmaking systems should keep working with full transparency. If people fully understand EBMM and start having a worse experience because they know what is going on then something is flawed in the process. It probably means it's "abusing" the user and ruining the experience the user actually wants to have when he's playing his videogame.

So full transparency seems to be the only way to avoid this becoming a terrible thing in the gaming industry. Because these matchmaking systems aren't necessarily bad (I like SBMM in Modern Warfare).

It reminds me of Google ads and tracking on the internet. All the secretism makes people shy away from the idea but if we really think about it, it's great stuff! You get better ads that are actually aimed at you and what you are interested in, ultimately giving you a better experience. But no transparency just makes it a creepy and abusive system.

Deleted member 16657 · Sep 20, 2020

I had no idea this was a thing and that devs were trojan horsing it in under the guise of skill based matchmaking. I guess we can assume every big budget multiplayer game is using some form of this?

Deleted member 8752 · Sep 20, 2020

Einchy said:
I hate all this type of shit, just find me a lobby with the best ping.

Well said. I used to get annihilated online every time I played. And it was fun because it made me better. Now, I have to play against other scrubs and it's harder to get better. Definitely prefer completely randomized match making.

collige · Sep 20, 2020

PogChamp said:
I had no idea this was a thing and that devs were trojan horsing it in under the guise of skill based matchmaking. I guess we can assume every big budget multiplayer game is using some form of this?

No because...

To our best knowledge, we have not seen any existing matchmaking method that formally treats matchmaking as an optimization problem to maximize player engagement

Paper is from 2017, it takes time to go from a theoretical paper -> test implementation in one game -> popularization

Deleted member 16657 · Sep 20, 2020

collige said:
No because...

Paper is from 2017, it takes time to go from a theoretical paper -> test implementation in one game -> popularization

Ah I see. I wonder how we could get information on this. Maybe tracking thousands of games would show if this was currently implemented

SapientWolf · Sep 20, 2020

I'm 99% sure CoD has a form of this. I think these systems can be more punishing for people that are on the far end of the bell curve in both directions, i.e. far above average and the far below average. You would definitely notice being moved up or down a tier at the extremes. The gaps between player skill would be wider because there are less players in those tiers.

Battlefield doesn't really need it because if you randomly throw 128 players together the skill levels tend to average out.

Hallucinations · Sep 20, 2020

The problem with these studies if often how long they go for.

From my limited knowledge on data and analytics in terms of content engagement with broader digital media (studying a post-grad in comms). A lot of the research only lasts a few weeks or months. A lot of it gets mitigated in the long term.

Maybe for a multiplayer game this is important early on. Maybe not?? Maybe just some key fundamentals should be utilised.

And maybe just ... focus on designing better maps, better guns and mechanics and being aware of general entertainment trends and what that means.

It doesn't necessarily require using expensive analytics software or testers or whatever position is going to come out of making this a thing.

Like do people think Minecraft got big because Notch understood anything about keeping players engaged early on? Doubt it considering its unexplained mechanics requiring you to dig through wiki's when the game was young.

No. It was the fact that YouTube and we live in a participatory culture via a Web 2.0 world.

collige · Sep 20, 2020

PogChamp said:
Ah I see. I wonder how we could get information on this. Maybe tracking thousands of games would show if this was currently implemented

Programmers love to talk about this stuff, I bet if it was successfully implemented there'd be a blog post or GDC talk about it. You need a good skill based matchmaking system in the first place for this which is a hard enough problem on its own.

Orb · Sep 20, 2020

People have been complaining about SBMM / EBMM in MW2019 since it launched. It's my second most played game of all time with hundreds of hours in it, and I just don't "feel" it the way that many people claim to. I don't deny that some level of it is there, I just don't really notice. I never feel like I'm being punished by the matchmaking system. So it doesn't bother me.

Alek · Sep 20, 2020

collige said:
No because...

Paper is from 2017, it takes time to go from a theoretical paper -> test implementation in one game -> popularization

That's not how a lot of things work inside industry. To be honest, it's rare that industry lead research is published at all, and from my personal experience, it's not at all uncommon for developers to speak about matchmaking as an engagement problem.

Often times, research is actually lagging behind what is actually being practiced, inside industry, because studios do not want to publish what they are doing, for fear of their competitors adopting the same strategies. For instance, Ubisoft published their SDT survey, describing how they were applying motivational models to improve player engagement in their games, in 2018, however that survey was first used on The Division in 2016, likely during development sometime between 2014 and 2015, 3-4 years prior to the publication of the research.

The authors also write 'to the best of our knowledge' they're unaware of any title that 'formally' treats the matchmaking system as an engagement problem. But the reality is, that unless you're working within any particular studio, 'the best of your knowledge' is very limited, everything is NDAed and studios often do not publicly share information that they might feel provides an advantage against their competitors.

Your thoughts on Engagement Based Matchmaking?

Games User Researcher

Games User Researcher

Games User Researcher

Games User Researcher

Attempted to circumvent ban with alt account

Games User Researcher

Alt Account

Games User Researcher

Games User Researcher

Prophet of Regret

User requested account closure

User requested account closure

User requested account closure

Games User Researcher