Skill, Matchmaking, and Ranking Systems Design

In this 2016 GDC sesion, Activision’s Josh Menke discusses how skill can be properly used for matchmaking and how good matchmaking can free gameplay designers to design with depth, and suggests best practices for ranking systems.

Register for GDC:

Join the GDC mailing list:

In this 2016 GDC sesion, Activision’s Josh Menke discusses how skill can be properly used for matchmaking and how good matchmaking can free gameplay designers to design with depth, and suggests best practices for ranking systems.

Follow GDC on Twitter:

GDC talks cover a range of developmental topics including game design, programming, audio, visual arts, business management, production, online games, and much more. We post a fresh GDC video every day. Subscribe to the channel to stay on top of regular updates, and check out GDC Vault for thousands of more in-depth talks from our archives.


Xem thêm bài viết khác:

  1. Mario D. Zmaj 9 months ago

    I wish there is an option to play a game without all this bullshit these overpriced n00b idiots implemented

  2. BLOOD WOLF 9 months ago

    Mechwarrior Online right now 23:15

  3. Giranam0 Harper 9 months ago

    It's always been interesting to me to watch Josh try to show angry people why they're wrong. He keeps pretty calm and just presents information more-so than opinion and people mistake it for opinion and not just presenting factual information (reporting, not interpreting). His experience is vast and deep and the information he's privy to is much greater than the angry players upset by the systems in place so when they get angry at him, all I really see is someone who's mad that the results of the game got documented in great mass, calculated, then put into play. Like if I had 10 kills and 20 deaths one game, when all people in the lobby typically average 15 kills and 10 deaths, then I get mad at the result and rather than trying to improve myself, I'd just blame the system for not making sure I was supposed to be the top player in the lobby.

  4. Arash 9 months ago

    Modern warfare 2019 sbmm is garbage btw

  5. Iamyour FAN 9 months ago

    Let one thing be said tough: The tighter your matchmaking, and the fairer your games, the less rewarding will actually being good at your game be.

    If i put a 1000 hours in your game and you will only matchmake me with other people of equal skill, i will never get the feeling of actually improving, since my opponents improve alongside me. This can be extremely frustrating and keep players from trying to master your game and put a long time into it.

    Matchmaking sound so obvious and necessary, but it definitely has it's downsides which must be considererd.

  6. Marcus 9 months ago

    The guy didn't go easy on the memes.

  7. Cubapete 9 months ago

    why not just use a server browser?

  8. PlanarHOTS 9 months ago

    Well, wrote a post on a certain forum detailing this talk, so here are the approx question times, as i have them:

    And the details of the questions:
    Q1: 42:30–44:20 about server sending lots of games versus sending one game to the client. also about player churn between these two.

    Q2: 43:25–46:00 have you ever been privy to any goals or motivation for not having a 50/50 win-loss goal?

    Q3: 46:05–47:20 The skil system you talked about based on win/losses, similar to elo, have you though about using more in game metrics to rank players, like about different roles and position in moba's?

    Q4: 47:25–48:30 experience on small player games–how large does the playerbase need to be to make a skill based matchmaking system make sense?

    Q5: 48:30–50:30 What are your thoughts on decay, about skill being lowered the longer someone does not play. such as if someone doesn't play a moba for a month the meta changes, even if the individual player's skill does not.

    Q6: 50:35–51:10 It seems like a player's skill variance would account for someone not playing for a while, so why decay the rating? Isn't the math supposed to handle that?

    Q7: 51:15–50:40 When you matchmake, you do it on the chunks right? Not granularly? Like if you're going to match anyone in silver, you just match them against anyone in silver–or do you do 2200 versus 2300?

    Q8: 50:40–52:25 The popularity of the hybrid system, I feel like LoL and these games are like going to go through this progression based system, and then transfer you over when we're sure you're ready, because we don't want you to drop or bounce because we're seeing this number go down. Can you talk a little bit more about that, because you were saying "oh, just go skill" which is obviously a seductive thing if you're making an esport kind of game, but you're just worried about people just bouncing from the game.

    Q9: 52:30–53:40 Do you have critical mass issues in games? how you weigh the tradeoffs between giving people matches that are not great for them to wait long, versus making them wait to get a better match, and how you balance those factors?

    Q10: 53:45–55:15 So for some games that have different roles, players might have different skill levels, depending on what they end up doing. Any thoughts on hwo you can matchmake and determine skill based on the roll they end up playing? both about situations where the role is pre-chosen and is post-chosen of mathmaking

    Q11: 55:20–56:50 What are the dynamics of skill over time? So if you start off with a given skill distribution, you showed high skill players at a constant skill, but bad players skill creeping up over time earlier. Do you see the opposite effect where everyone is normally distributed that the variance of skill narrows, or feels constant over a game?

    Q12: 56:55–58:25 How do you think ranked reset effect hybrid models? Once a year you reset, and good players get a higher rank to start, and lower players a lower rank. For example in HS, players reset every month. Players get caught up in the middle after reset and causing discomfort in bad players.

    Q13: 58:30–59:25 But do you think maybe you would have, just a larger difference in reset? Say if i was a 11 player in HS who resets to rank 15, if i set them at rank 12 instead, where they're unlikely to meet up with newer players, or bad players, do you think there are different ways like that?

    Q14: 59:30–1:00:40 How would you deal with players who might intentionally lose to "smurf" to try to play against worse players?

  9. JtheDruggernaut 9 months ago

    Daddy Menke is lit af

  10. Cyrus Cuenca 9 months ago

    This was really helpful. Thank you!

  11. Kristopher Driver 9 months ago

    Monster hunter had all of these gripes including the hybrid ranking system wall lol

  12. J-Ryze 9 months ago

    Please let this man work on League.

  13. _Masko_ 9 months ago

    I don't wanna sound toxic, but looking at the comment section… It feels like most of the people here are way to dumb to understand what Josh is trying to explain.
    Maybe read a book instead of reading Twitter and salty Reddit posts.

  14. NewSystem 9 months ago

    just put the player together and let them kill eachother, for god sakes

  15. Jakub Pawlowski 9 months ago

    Although I disagree with many opinions presented I still thank you very much for sharing. It helped me ensure I’m not missing anything and made me more confident that I’m headed the right direction.

  16. Basu 9 months ago

    The design philosophies he preaches here and the reality of Halo 5's matchmaking could not be further apart.

  17. daz 9 months ago

    this was an amazing discussion, sucks that some of the top comments are focusing on something not even related

  18. Luke TheNotable 9 months ago

    How much dedicated RAM should my minecraft server have?

  19. Southern Hemisphere 9 months ago

    When will there be an interview on these game companies web site designs? Is there already a video on a discussion like this?

  20. Gabriel Stav 9 months ago

    Great talk and interesting to think about how this works in OW.

  21. DEvineSz 〈3 9 months ago

    sigh… 29 minutes in, again did he just say if he has an already good progression system in-game which is not a competitive match making system he would rather not put a competitive match making system because it is like the other progression system?, are you kidding me? having a single or even multi player progression system is nothing like a competitive mm system, those systems are good for players who want to do a specific thing match their skill to someone else and gain skill progression in a certain task in-game which usually is PVP.
    having a single or multi progression systems that is not match making is in no way similar to this maybe in ur "idea of the game" mind it is but trust me not for players who play completely different parts of your game, some like pve, some pvp, some like both.

  22. monnef 9 months ago

    Really, banning smurfs? Isn't a better solution to quickly adjust rating to match skill? I can't image I would buy again Overwatch to play a different "main" in competitive (since teammates see your stats, you can't really play anything else, without being reported as toxic/trolling/throwing) and be banned after a first game…

  23. Evigmae VonBenerot 9 months ago

    I feel like these systems never really work on team games. I get 1v1 might be easily handled by the math. But 5v5s have too many abstracts apparently. However, I feel Machine Learning could easily solve all of this. Either that, or actually let users influence the systems with votes after each match.

  24. Teneban 9 months ago

    For the game I'm currently making, I'm trying something "not so new". A difficulty slider. We have had those in PvE for decades, but I think they would fit great in PvP.
    The idea is, people who play in "easy" mode are primarily matched against people lower than themselves who are playing in "hard" mode – but the key is that they choose the experience they want to get, every single time.
    With that system, I'm able to have some players with a 65% win rate in bronze, and players with a 35% win rate in platinum. The goal of the system is not to provide an even match anymore – it's to provide the experience the player is looking for.

    Sometimes, a hardcore player may want to take it easy. For example, maybe they're testing a new character, and they expect to get sub-par results with that new character for a while.
    Maybe they're super engaged community figures, who want to play against their fans and give them precious pieces of advice when the game ends.
    Maybe they're top players who don't want to wait too long because if they were looking for an even match, they would just ask their top player friends, not the match making server.
    Or maybe they're just having a bad day and want to stomp over new players.
    And there can be just as many reasons why a casual player might want to play in hard mode.
    Due to the nature of the ranking system, players playing easy mode wouldn't be rewarded too much for winning, and players playing hard mode wouldn't be punished too much for losing.
    But either way, by asking the player "what are you looking for", I'm able to provide that experience.

    PS: about Elo – the main issue the presenter seems to have with it is that it's slow… But that's what K-values are for, aren't they? Did I miss something that makes playing with K-values bad for some reason?

  25. absol_89 9 months ago

    I tried making a skill rating algorithm for any scoreboard based team game. Tried it on Battlefield 1 and it turned out well 😀

  26. DEvineSz 〈3 9 months ago

    16 minutes in, did he just say it's OK to give player that wait longer a worse match-up?????? did I understand him correctly? you should not give players that wait a long time a match at all if u can't find one you should send them a message that says there are currently no match ups that fit your skill and ASK them would you like to search for a lesser match? then if they wanna take a risk of wasting their time with bad players it's atleast their choice to do so, cuz trust me in some games i'd rather be on queue and minimize the game do something else than waste my time with bad players.

  27. DEvineSz 〈3 9 months ago

    11 minutes in, Just have to say something, Players that play games "seriously" meaning they get invested in them would rather wait to find a good match-up than find a fast match that would end in either direction really, meaning I as a good player in x game would rather wait than destroy another player because he does not match my skill and ranking and surely he will rather wait too. so in the end of the day the player that wins a lot will get frustrated too.

  28. Ersen 9 months ago

    I know this is from 2016, but it's unfortunate that you got someone from Activision of all companies to talk about matchmaking. Especially sharing it now, after we found out about their plans on what to do with matchmaking.

  29. jort93z 9 months ago

    Very interesting talk

  30. maggit 9 months ago

    This coming from a guy whose parent company wants to introduce microtransaction based matchmaking.

  31. LazärheaD 9 months ago

    Or you can put in a system that convinces people in multiplayer games to purchase items for a game through microtransactions, when matchmaked against people that purchased microtransaction weapons already.

Leave a Comment

Your email address will not be published.

You may also like