Rioter Comments
patmax17 (EUW)
: First of all, congratulations for listening to your players like this. Personally I have no interest in SoloQ, but I hope this will at least calm down some of the haters One question about the new queue: How do you think this will impact queue times of 4-men premades? They need a solo player queued in dynamic to start a match (actually, 98% of the cases two single players queued in dynamic will be needed), do you think there will still be enough of them?
> [{quoted}](name=patmax17,realm=EUW,application-id=cIfEodbz,discussion-id=HGblb9fH,comment-id=0025,timestamp=2016-01-19T08:40:38.027+0000) > > First of all, congratulations for listening to your players like this. Personally I have no interest in SoloQ, but I hope this will at least calm down some of the haters > > One question about the new queue: How do you think this will impact queue times of 4-men premades? They need a solo player queued in dynamic to start a match (actually, 98% of the cases two single players queued in dynamic will be needed), do you think there will still be enough of them? To be honest, we'll just have to see and keep a close eye on it. It's too hard to predict right now without some live data.
AlexdInsane (EUNE)
: In the solo ranked ladder are we going to be able to play duo with a friend, or its solo and only solo?
> [{quoted}](name=AlexdInsane,realm=EUNE,application-id=cIfEodbz,discussion-id=HGblb9fH,comment-id=0028,timestamp=2016-01-19T08:51:02.700+0000) > > In the solo ranked ladder are we going to be able to play duo with a friend, or its solo and only solo? It's a pure Solo Queue, so only solos.
: Thanks for listening! Question if I may: will our initial MMR for Dynamic Queue be affected by our Season 6 Solo Queue placements / matches, or only our preseason MMR?
> [{quoted}](name=SuperMrBlob,realm=OCE,application-id=cIfEodbz,discussion-id=HGblb9fH,comment-id=002d,timestamp=2016-01-19T08:55:30.540+0000) > > Thanks for listening! > > Question if I may: will our initial MMR for Dynamic Queue be affected by our Season 6 Solo Queue placements / matches, or only our preseason MMR? Initial MMR for Dynamic Queues will be affected by your Season 5 Ranked Solo/Duo Queue rating. If Solo Queue launches and you've played Solo Queue games but have NOT played Dynamic Queues yet, then your initial MMR for Dynamic Queues will now be affected by your current Season 6 Solo Queue rating.
Ludovico (EUW)
: Wait a second... if there´s no team-ranked in S6, what about TT ranked?
> [{quoted}](name=Ludovico,realm=EUW,application-id=cIfEodbz,discussion-id=HGblb9fH,comment-id=002e,timestamp=2016-01-19T08:56:25.282+0000) > > Wait a second... if there´s no team-ranked in S6, what about TT ranked? TT Ranked is unaffected at this time.
Shallers (NA)
: Does this mean separate ladders and MMR's for the two separate queues as well?
Separate ladders and MMRs; however, performance in Solo Queue will influence your initial ratings in Dynamic Queue. Performance in Dynamic Queue will NOT affect your ratings in Solo Queue.
Aniion (NA)
: >Expect most of the Ranked Rewards to shift to Dynamic Ranked I know you haven't decided yet, but would that mean just queuing up in Dynamic Ranked, or would it be based on party size? E.g. some sort of separate progression within Dynamic Ranked where your party size is included in which rewards you get? I'm asking because I'd rather get more/better ranked rewards and deal with premades, but I'm still a solo player. If playing solo in Dynamic is the same as playing solo in Solo Ranked, than there's no real point in going Dynamic Ranked. But if I would be in the running for better/more Ranked Rewards, I'd queue Dynamic more.
It just means queuing into Dynamic Queue. There will be 2 options in the queue selector, and you can queue up solo in Dynamic Queue for a chance at the rewards. We're still figuring out how rewards work, but that's the gist of it.
Rotakn (NA)
: Will you be able to queue in the Dynamic Ranked as solo still?
  Rioter Comments
: > Last year, we investigated why players didn’t play ranked, and a huge number said they didn’t want to abandon their friends to do it. In other words, there was a large subset of players who wanted to play competitively with friends on a ranked ladder, but lacked a great avenue to do so. I don't have access to your data or your survey methodology, but I believe it's possible you misinterpreted your user research. Perhaps those players would prefer to play ranked with their friends, but they could have easily formed a ranked team, so you have to ask yourself why they weren't. Dynamic queue seems to solve for the objection that it's hard to get 5 friends on at the same time, but is that the main objection? Could it be that individual ranking & pride in individual accomplishment, by its intrinsic nature, makes it very challenging to play with people who you want to get along with? Could it be that team environments with aligned individual incentives to be good teammates and friends are best served by recognizing the accomplishments of the team as the primary metric for success? I believe that's the case. I don't want to play solo-rank with my friends because we have different skill levels. I don't play much ranked because I don't want to abandon my friends, but I don't want ranked to change so they can play with me. I think Riot's core business revolves around people believing they could be at the top. Dynamic queue is a tradeoff between competitiveness and fun. If your business goals align with making the game more fun, then this may be a move in the right direction. But I think making the game more fun is not necessarily in your best interests as a business. That is, not when the fun has a trade off. Broad live testing is certainly the right approach, but be careful. Even if this game is merely perceived to be less competitive (and that perception is wrong), it may be enough to damage the real value in your business. Riot isn't so successful because they have the most fun game on the market. Fun isn't enough. Don't lose sight of that.
The study was pretty extensive last year, and DID include entire sections on WHY players prefer to play Ranked with their friends, but couldn't actually make a Ranked Team. In fact, as you suggest, most of the reasons had to do with high barriers of entry to Ranked Teams, requiring a small roster of 5 players to be online at the same time, and Ranked Team roster sizes. There were also issues with queue times and match quality because there weren't enough teams in Ranked Teams to create good games. Remember that even in Dynamic Queues, you still have Tier Restrictions in place. You are NOT going to be able to rank with your friends if they are different skill levels, even if you wanted to. Tier Restrictions are one of the primary reasons why competitive integrity is relatively persistent in the Dynamic Queue system.
Whober (NA)
: > [{quoted}](name=Lyte,realm=NA,application-id=A7LBtoKc,discussion-id=XvqbAgrF,comment-id=,timestamp=2016-01-16T05:48:06.262+0000) > * **When we polled players who were playing in dynamic queues the specific question**: “I like that you are allowed to queue with any number of people to play ranked”, about 55% of players agreed, 20% were neutral, and 25% disagreed. We’re not saying this to sway opinions either way--it tells us the system is both polarizing, and broadly desired. We’ll check in on this statistic again next week. The issue with this people who don't feel like playing against fully stacked teams won't be even queueing up for dynamic queue, so you won't be able to get a good poll unless you poll players playing ALL game modes.
The surveys were sent to a random sample of the entire playerbase. You can then take the surveys and segment the demographics to run further analyses. This wasn't clear in the original post, so I've corrected it.
: >We hear you on the impact dynamic queues have on the measurement of solo skill. We’re currently doing everything we can to make dynamic queues a great experience, and we always had back up plans for the 2016 ranked season. The way Dynamic queue has been introduced on the [2016 season ranked promo site](http://na.leagueoflegends.com/en/site/2016-season-update/ranked-improvements.html) does not give the impression that there is any possibility of anything other than the replacement of Solo queue with Dynamic queue happening. There is no other way to interpret >This upcoming season we're replacing the solo/duo queue with a dynamic group queue, as you doing anything other than replacing solo queue with dynamic queue. So it strikes me as really dishonest when [on Twitter you imply](https://twitter.com/RiotLyte/status/688070169290125315) that there is the possibility of solo queue staying when the official material does not leave that as a possibility.
The big promo website talks about a huge number of features and systems, and you know what? Things are going to change. Not every feature or system in that website is going to launch as described in the website because every team is collecting data, learning lessons, and trying new things. You aren't going to see a website like that say, "THIS IS AN EXPERIMENTAL FEATURE THAT MAY OR MAY NOT CHANGE IN ITS FINAL FORM" or else we'd have to plaster that on every feature and system article. The website information is not more "official" than information straight from the team working on the product. Even 2 months ago, we mentioned that Dynamic Queues were a test that we weren't sure would work out.
: What I would like you to do, Lyte, is explain why they don't want to include a true "solo queue," where it is nothing but five solos playing against five other solos. There has been such a large outcry, surely you must know this. Presumably, this is because you want groups of 4 to be able to play, and there would be too few solo players to fill in the gaps because they would probably all go solo, right? If so, this leads me to a couple of questions. 1) What percentage of people are playing solo now compared to the percentage that are in groups? 1a) Is this percentage too small to have their wishes be considered? Why or why not? 2) Will there actually be a shortage of people going solo in dynamic queue, if a solo queue is implemented? (Yes, no projections will match actually implementing the change live, etc. Give your best educated guess.) 2a) If yes, would it not be an option to perhaps give some incentive for solo players to sign up for dynamic queue to fix this shortage? (I could think of a few different reward systems, but I'm not arrogant enough to presume that I would know better than the whole of Riot would.) 3) You mention that you want premade 4s/5s to match up 95% of the time (which is much too small a percentage, btw). What about other team makeup percentages? E.g. 1v1+1+3, 2v1+1+1+1+1, 3+1+1v3+2, etc. etc. 3a) Will you be keeping all other makeups as loosely arranged as you are with 4s/5s? tl;dr--Lyte, please explain with data why there cannot be a true solo queue alongside the dynamic queue. Thank you for your transparency. Keep it up.
One of the biggest issues with a true Solo Queue is playerbase segmentation. Players in NA don't know this, but many LoL servers around the world don't even have all queues; in fact, some servers only have Ranked Solo/Duo Queue for specific times of the day, or specific days of the week, and no Twisted Treeline, Normal Draft or Ranked Teams at all. So for these servers, it would have been a HUGE benefit to players if we could have a single competitive queue instead of forcing players to either play solo, or play with friends in a ranked environment. This is why we wanted to see if Dynamic Queues could work, and we are doing everything we can to see if we can make it a great experience. If it doesn't work, it doesn't work, and we have other ideas in mind. We know it's hard to believe, but there are more players interested in playing a competitive ranked queue with friends than there are players interested in a solo ranked queue experience. We still don't know what's the best final call and we're working non-stop looking at the data this weekend, but these decisions aren't as easy as some players might think.
: Lyte, you're the one who was saying that vocal chat in game will raise harassement and toxicity without even trying to test it on LoL. But with the Dynamic Queue you give to premade, the possibility to outplay any solo player, not with something in the game, but with something outside, you are giving an edge to people in a competitive environment. Look at you differents concurrent : - CSGO : Dynamic Queue in Ranked but Voice Chat for everyone, premade don't have an advantage over solo player. - Dota 2 : 3 differents Ladder : Solo (1 player) => Solo Elo/MMR, Party (2 and 3 players) => Party Elo/MMR and you can play with any friend, you have a different ELO that Solo and Team (5 players). As you can see, there is no 4+1 or 3+1+1, because again, giving an advantage in normal isn't a problem, in the end there is no Ladder, but giving an advantage in a competitive mode isn't logical and **fair** at all. HOTS : Was using Dynamic Queue, all the community was against, they changed it and take exemple on you and 2 days ago they annonced that they will do something quasi-similar in normal game. No one want to be the solo player with 4 premade, because the solo players knows that their decision will not be taken in consideration during the game. Seriously, think about it, the moment a premade take a decision with vocal and do it the solo players will not have one second to say something, because the team will already be on the move. Instead of being one of the Actor of the game, solo player will become just an Extra, following the hero and his friend on their adventure.
To be clear, I think you're making a few assumptions. It's pretty universally agreed among industry experts that voice chat introduces some amount of toxicity into an ecosystem or game; however, that doesn't mean we are opposed to it. We definitely understand the importance of communication in a competitive team-based game, and every time we talk about voice chat, we mention that we're open to it, we'd just want to change some of the basic designs of voice chat to mitigate some of the known issues.
: Wish they would actually release the data, or create more balanced questions for when they asked. FOR EXAMPLE: If instead of asking *“I like that you are allowed to queue with any number of people to play ranked”* you should have asked *“I like that you are allowed to queue with any number of people to play ranked in the same queue as Solo-queue."* When you load the question in your favor, of course people will agree. FURTHERMORE: If you polled people ALREADY IN THE DYNAMIC QUE, and they are testing Dynamic Queue, then they are almost certainly queuing up with friends, and will then support it. You literally polled a group of people that was already in the system.
Multiple scientists worked on the surveys, and actually the same question was asked 5 different ways if I remember. Slight changes in wording, different frames of reference, they asked it all, and the results all trended the same direction. The survey tool they used also assesses a random sample of the entire playerbase, not just players currently in the system.
  Rioter Comments
  Rioter Comments
: Lyte has been permantely banned from Boards
: Just out of curiosity why do you do all your testing on the NA servers which has your largest server population instead of testing it out on smaller servers that would effect less people if something goes wrong first?
> [{quoted}](name=Fongletto,realm=OCE,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=001000000004,timestamp=2015-11-12T22:13:09.212+0000) > > Just out of curiosity why do you do all your testing on the NA servers which has your largest server population instead of testing it out on smaller servers that would effect less people if something goes wrong first? We see this question a lot, so going to jump in even though it's late in the thread. When we choose to test new systems and features, it has everything to do with timezone. Although some teams work in multiple offices around the world, 100% of our team works in the LA office which means we prefer to choose a test server that has a timezone that does **NOT** completely wreck our sleep schedules. You generally want to turn on a new feature or system during off-peak hours, say 6 AM, and then monitor it for about 24 hours at a time. By choosing NA, we can basically wake up and sleep at an OK time which is far less exhausting for the team.
Leiada (NA)
: I was on the PBE and I remember that people were dodging all the time because it wasn't clear that they had to lock everything in. Will it be clearer in the future?
> [{quoted}](name=TheUnkindRaven,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0043,timestamp=2015-11-12T20:03:35.195+0000) > > I was on the PBE and I remember that people were dodging all the time because it wasn't clear that they had to lock everything in. Will it be clearer in the future? Yes, the latest version internally has a few more features that clarify that you **MUST** lock in all Bans and Picks. We'll keep an eye on this as we go, we know this was not clear enough on the PBE release for sure.
: I like all of these new changes, but I cant help but feel like the new champ select will limit the roles that people know how to play. For example, in solo que as it is right now, you have to be ready to play any role, as you are not guaranteed a high pick, and sometimes its best to fill because you have teammates calling mid only or feed etc. With the way things are right now, you might get your role, you might not, and after a few 100 games you understand just about every role in the game, and I think that in turn makes you a much stronger player. I find a top laner who understands the hardships of a jnglr plays better when getting ganked, asking for ganks, or an adc plays better when he knows how supporting generally works or vice versa. I main adc and support and based on my experience the "adc only can't support" guy generally never plays well because he is lacking knowledge in half of what is going on in bot lane. With these new changes, if everyone can just play the role they want, not only am I afraid it will increase que times like in team builder(which i feel like has been addressed somewhere but I never saw it), I am also afraid players will only have experience with 2 maybe 3 roles at most. We would all love to only play mid or adc or whatever we main and whatever our secondary position is, but by being forced to learn all of the roles I think we become better players who understand more aspects of the game and can better support our teammates because we have more general knowledge in what is going on. Like I said I main bot lane roles, but having played my fair share of top lane and jungling, I definitely can be more understanding throughout the game because I know the role almost as well as I know adc, even though I would never select top lane in the new champ select.
> [{quoted}](name=Bulletoverload,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0030,timestamp=2015-11-12T19:47:41.058+0000) > > I like all of these new changes, but I cant help but feel like the new champ select will limit the roles that people know how to play. For example, in solo que as it is right now, you have to be ready to play any role, as you are not guaranteed a high pick, and sometimes its best to fill because you have teammates calling mid only or feed etc. With the way things are right now, you might get your role, you might not, and after a few 100 games you understand just about every role in the game, and I think that in turn makes you a much stronger player. > > I find a top laner who understands the hardships of a jnglr plays better when getting ganked, asking for ganks, or an adc plays better when he knows how supporting generally works or vice versa. I main adc and support and based on my experience the "adc only can't support" guy generally never plays well because he is lacking knowledge in half of what is going on in bot lane. With these new changes, if everyone can just play the role they want, not only am I afraid it will increase que times like in team builder(which i feel like has been addressed somewhere but I never saw it), I am also afraid players will only have experience with 2 maybe 3 roles at most. > > We would all love to only play mid or adc or whatever we main and whatever our secondary position is, but by being forced to learn all of the roles I think we become better players who understand more aspects of the game and can better support our teammates because we have more general knowledge in what is going on. Like I said I main bot lane roles, but having played my fair share of top lane and jungling, I definitely can be more understanding throughout the game because I know the role almost as well as I know adc, even though I would never select top lane in the new champ select. This is a pretty common misunderstanding in Ranked play today. On average, players are only proficient at about 2.2 positions today. In fact, as you go higher in skill, you become more specialized in particular positions. However, we have some ideas to actually encourage players to master multiple positions beyond 2, and potentially reward players for truly being a master of all positions in League.
Rismosch (EUW)
: ***
> [{quoted}](name=Rismosch,realm=EUW,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0063,timestamp=2015-11-12T20:32:56.486+0000) > > Team Builder will be kind of useless, because now with the new draft you get the role that you want to play. What will happen to Team Builder after the release of the new draft? Are you going to change something or do you remove it completely? All the testing so far has suggested that the new Champ Select is the best possible League experience, so we're considering sunsetting the old Team Builder. We'd also like to replace some of the other queue experiences (like Normal Draft and Ranked) with this new Champ Select experience.
: I am not 100% sure if this is a 'huge' issue. However, in the old champ select, if you bought a skin during the pick phase, you wouldn't be able to select it that game. Will this be an issue in the new champ select? Thanks, See yah on the rift! {{summoner:4}} {{summoner:4}} {{summoner:4}} {{sticker:slayer-jinx-catface}}
> [{quoted}](name=Aeronomical,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0051,timestamp=2015-11-12T20:15:54.331+0000) > > I am not 100% sure if this is a 'huge' issue. However, in the old champ select, if you bought a skin during the pick phase, you wouldn't be able to select it that game. Will this be an issue in the new champ select? > Thanks, > See yah on the rift! {{summoner:4}} {{summoner:4}} {{summoner:4}} > {{sticker:slayer-jinx-catface}} We're going to try to make sure when you buy the Skin in Champ Select, you can use it that game. So when 4 of your teammates have PROJECT skins... you know what to do :)
: So, first of all I want to say that you created something awesome there. It looks really nice and the banphase is -in my opinion- really great so there is no possibility of the first pick, may he be a troll, trollbanning three champs. Now coming to my questions: 1.) Why did you choose the last three people to ban instead of the first three? 2.) What if a summoner, who picked 'fill' as one of his roles/positions gets to fill a position he actually doesn't want to play? Will there be a chance to kick this one in champselect and return to queue? Or will there be something like an extra reason to report him after the game? And what will you do against people like this?
> [{quoted}](name=VV ShockwavE,realm=EUW,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=002d,timestamp=2015-11-12T19:44:45.437+0000) > > So, first of all I want to say that you created something awesome there. It looks really nice and the banphase is -in my opinion- really great so there is no possibility of the first pick, may he be a troll, trollbanning three champs. > > Now coming to my questions: > > 1.) Why did you choose the last three people to ban instead of the first three? > 2.) What if a summoner, who picked 'fill' as one of his roles/positions gets to fill a position he actually doesn't want to play? Will there be a chance to kick this one in champselect and return to queue? Or will there be something like an extra reason to report him after the game? And what will you do against people like this? boourns remembers the crazy testing we had to do with Distributed Bans. At the beginning, we actually had Players 1, 2, 3, get the bans, then Players 4 and 5 get 1st and 2nd pick. However, this was super confusing to players in the testing because they already had **years** of prior experience of picks "starting" from the top of the team. So, we swapped it the current format where Players 3, 4, 5 get the bans, then we "reset" into Pick Phase and start from the top. This was much clearer for current players of the game, so we stuck with it.
EHMG (NA)
: The new way seems great and useful for the current meta. But I think the meta might be shifted a bit with the new objective introduced for top Laners , and the new masteries , especially Vampirism, which can give some sustain. May be adc and support would prefer to begin top lane now and I would like the new champion select to be able to go with the meta shift if it proves useful
> [{quoted}](name=EHMG,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0034,timestamp=2015-11-12T19:49:51.626+0000) > > The new way seems great and useful for the current meta. > But I think the meta might be shifted a bit with the new objective introduced for top Laners , and the new masteries , especially Vampirism, which can give some sustain. May be adc and support would prefer to begin top lane now and I would like the new champion select to be able to go with the meta shift if it proves useful The cool part about the new Champ Select system is how flexible it is. If the meta changes to 2:1:1:J for example, we can literally input that into the matchmaker and start having the system create that team comp. If there is a day where multiple team comps are popular, we have some backup plans to allow teams to opt-in to which comps they are interested in playing.
: Will it be possible to report the champion select trollers in anyway?
> [{quoted}](name=Flammenklinge77,realm=EUW,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0035,timestamp=2015-11-12T19:51:01.586+0000) > > Will it be possible to report the champion select trollers in anyway? There won't be any additional features for Champ Select trolling or votekicking initially. We think the design resolves a lot of the current conflicts of Champ Select, but are keeping an eye on any remaining issues. If there are still significant problems in Champ Select, we may consider a way for players to queue dodge and file a report and a reason for why they dodged, and if we can do some machine learning magic here we'll make sure the players don't get a queue dodge penalty if the report was accurate.
: I have always wanted to play with a rioter .3.
> [{quoted}](name=TubbyLerue,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=00370000,timestamp=2015-11-12T20:10:17.918+0000) > > I have always wanted to play with a rioter .3. Maybe we should setup a community game. Dev team versus the world.
Tuhglife (EUW)
: i like the design and how you can pick now your lane. but showing the champions before banning..... trolls will use it.
> [{quoted}](name=Tuhglife,realm=EUW,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=003f,timestamp=2015-11-12T19:59:21.475+0000) > > i like the design and how you can pick now your lane. but showing the champions before banning..... trolls will use it. Actually, we saw something pretty weird in our playtests with Distributed Bans and Pick Intent. Because every single player now has a powerful "choice" in Champ Select (whether it's a ban, or the top 2 picks), it actually **reduced** trolling of Pick Intent. So basically, players were far less likely to ban a Pick Intent (unless they thought it was a S-tier champion or something) because they knew that if they pissed off their teammates, their teammates could reverse-troll them with their powerful choice. If you're super worried about trolling in Pick Intent either way, you can always let the timer run out and not choose any champs at all and display a blank helmet icon.
: If you're planning on giving IP bonuses to people who play in-demand roles: A) Is this the only thing you're giving them? B) Will people who get put into that role via selecting 'fill' also get the IP bonuses? C) Is this your only plan to help alleviate the issues that normal team builder had with queue times?
> [{quoted}](name=Hellioning,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0024,timestamp=2015-11-12T19:38:42.452+0000) > > If you're planning on giving IP bonuses to people who play in-demand roles: > > A) Is this the only thing you're giving them? > B) Will people who get put into that role via selecting 'fill' also get the IP bonuses? > C) Is this your only plan to help alleviate the issues that normal team builder had with queue times? a) It really depends what the issues are when we launch. We have a couple different options including some bonus IP, some components for the future Crafting system, etc. b) Yes, players that select "Fill" will always get bonuses. c) We have a few backup plans that have to do with tweaking the matchmaker.
: Can I have my random button back, please?
> [{quoted}](name=Hellioning,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0029,timestamp=2015-11-12T19:41:44.085+0000) > > Can I have my random button back, please? We're going to get it back for Normal Draft, but it's going to still be removed in Ranked. The PBE was considered the "Ranked" experience, so we didn't add the button for now.
Pryotra (NA)
: Well, when is it going live? Even a rough estimate would be nice. Also, will people be able to see your champion mastery in select? For example, I'm lvl 5 mastery with Heimer. Would other players be able to see that this guy who wants to take Heimer in knows how to play him? Or would it be hidden until the loading screen?
> [{quoted}](name=Pryotra,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0008,timestamp=2015-11-12T18:55:09.741+0000) > > Well, when is it going live? Even a rough estimate would be nice. > > Also, will people be able to see your champion mastery in select? For example, I'm lvl 5 mastery with Heimer. Would other players be able to see that this guy who wants to take Heimer in knows how to play him? Or would it be hidden until the loading screen? To be honest, we're not sure when we're going to go live yet. We have some polish we'd like to do, maybe a new feature or two, and have one nasty bug we need to fix before the Ranked 2016 Season. The team would love to do another PBE test, and potentially a roll out to NA and TR before the Ranked 2016 Season begins to iron out any further kinks, but we're working around the clock right now and it's hard to predict when each of these pieces will finish. Right now, there are no plans to allow everyone to see your Champion Mastery in Champ Select. We may re-visit this in the future though, and would just have to balance out the negative behaviors that were created. For example, in a distant future when we have more than 5 levels of Champion Mastery, maybe we only show Champion Mastery **only** when you are Level 10 or higher--this is a significant investment, and something "cool" to show you're a master with a champion. However, Level 10 might be a lofty, challenging goal. If we decided to show, say Level 2 Champ Mastery or higher, it would create an immediate perception that if you WEREN'T showing Champ Mastery with a champ, that you shouldn't play Ranked at all. We're aware some players would love this type of "gate" on who you can play in Ranked, but it's not something we're going to consider today because there are more important issues to improve in Ranked Mode.
Siachi (NA)
: So is this something of an evolution of Team Builder, or would it be its own queue?
> [{quoted}](name=Siachi,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0004,timestamp=2015-11-12T18:45:39.723+0000) > > So is this something of an evolution of Team Builder, or would it be its own queue? When we started this project, we always felt like this was the "end-game" of Team Builder. We're actually considering closing down the old Team Builder queue, and slowly replacing many of the queues with this new Champ Select experience. So, the Ranked 2016 Season will just have this new Champ Select experience as its queue, and in the future, Normal Draft will just be this new Champ Select experience.
: If someone still decides to dodge because they don't get their primary choosen role how far back will the others go in select? Back to quene? Or simply back to the map that shows where everyone is going to go and just a new person pops into the dodged role?
> [{quoted}](name=xBeautifulEvilx,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0012,timestamp=2015-11-12T19:18:22.653+0000) > > If someone still decides to dodge because they don't get their primary choosen role how far back will the others go in select? Back to quene? Or simply back to the map that shows where everyone is going to go and just a new person pops into the dodged role? If a player dodges, they will get the same queue dodge penalty as today (5/15 minutes in Normals, more in Ranked and -LP). For the rest of the team, they will go back into the matchmaking queue searching for new opponents. Once opponents are found, it'll replay the map animation with the flags dropping and enter a new Champ Select.
Solash (EUW)
: So not to create risky expectations or anything, but is the plan to push the new champion select experience to live WITH the start of the 2016 season? Or perhaps slightly before?
> [{quoted}](name=Solash,realm=EUW,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0015,timestamp=2015-11-12T19:21:19.266+0000) > > So not to create risky expectations or anything, but is the plan to push the new champion select experience to live WITH the start of the 2016 season? Or perhaps slightly before? The honest answer is, we're not sure. The team's been working around the clock because of this one nasty bug in Champ Select that crashes some clients, and we recently identified the root cause but it's a super tricky situation. We'd love to do another PBE beta of the new Champ Select, and potentially roll it out in Normal Draft on some servers before the start of the 2016 Ranked Season; however, no promises because we're not sure when we can nail down this bug and finish some of the remaining features we'd like to do.
Thales (NA)
: Hi, Lyte! I tried out the new champion select on PBE last week. It looks really, really cool. Are you at all concerned that the huge visual change may confound attempts to gauge the actual impact of the functional changes? I like the choice to have the champion select quote play globally when champs get locked in. What prompted this change? Is there any concern about champ select quotes (Sona, etc.) that don't make much sense this way?
> [{quoted}](name=Thales,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=000b,timestamp=2015-11-12T19:07:43.774+0000) > > Hi, Lyte! I tried out the new champion select on PBE last week. > > It looks really, really cool. Are you at all concerned that the huge visual change may confound attempts to gauge the actual impact of the functional changes? > > I like the choice to have the champion select quote play globally when champs get locked in. What prompted this change? Is there any concern about champ select quotes (Sona, etc.) that don't make much sense this way? We're planning some deep dives and behind the scenes looks at the building of the new Champ Select, and our initial testing was actually with these ugly (sorry boourns!) prototypes that just had grey boxes and words. This is called UX lab testing in the games industry, and allows design teams to focus feedback and testing on the actual _functional experience_ of the system and not be distracted by the polish of fancy visuals or animations. It's only until pretty far in the project until you start adding in the stuff that makes the system look like magic. The Champ Select quotes being played was actually a spontaneous idea by RiotBahamut. He was tinkering with the Champ Select and thought it would be cool, and the team instantly loved how it felt. We're aware there could be some trolling or spammy experiences, but we think the overall value would be worthwhile. In fact, we'd love to add some more easter eggs to the system...
: > If a player fails to lock in a pick or ban, champ select ends and players return to queue Does this mean, it's impossible to wait until timer runs out to NOT ban anyone? Will we get option to abstain from banning or to ban random champion? I actually prefer banless games (even ranked) just because of challenge it provides.
> [{quoted}](name=OKIGorgon,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0013,timestamp=2015-11-12T19:18:31.844+0000) > > Does this mean, it's impossible to wait until timer runs out to NOT ban anyone? Will we get option to abstain from banning or to ban random champion? I actually prefer banless games (even ranked) just because of challenge it provides. No, we've intentionally removed the ability to let the timer run out to lock in a Pick, or skip a Ban. When you're in a team with strangers, skipping a ban is a confusing and frustrating experience for the rest of the team (and they often misunderstand it as the Captain being AFK). If you really want the challenge of skipping bans, grab 4 friends and join Dynamic Queue as a Premade of 5 and ban random champions.
GPU (EUW)
: Hey guys, when will this go live in EUW?
> [{quoted}](name=Tryiaxle,realm=EUW,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0010,timestamp=2015-11-12T19:18:13.841+0000) > > Hey guys, when will this go live in EUW? We'd like to do some pre-season testing on NA and TR first, and if things are going well, we'll rapidly roll it out globally.
: First, thank you for having these Q&As, I love hearing about your perspective on the changes you guys bring to the game! Do you have any estimates on your chance of getting your primary role vs secondary? Cuz it seems that, for example, if your primary was the highly contested mid lane and your secondary the underappreciated support of even fill, it seems that you might actually get placed in your secondary role more frequently than your primary. Is this a possibility? Are you also considering adding exceptions for the fill position? Personally, I feel comfortable on all the lanes, but I am an atrocious jungler. Is there anyway I could fill for any position that isn't jungle, or will I just have to choose 2 laning roles? Also, one final, more complicated question. How, if at all, do you think the new process will alter the pick/ban phase? Thanks for your time! I'm loving his the changes are looking so far!
> [{quoted}](name=CCdash2224,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=000e,timestamp=2015-11-12T19:13:10.564+0000) > > First, thank you for having these Q&As, I love hearing about your perspective on the changes you guys bring to the game! > > Do you have any estimates on your chance of getting your primary role vs secondary? Cuz it seems that, for example, if your primary was the highly contested mid lane and your secondary the underappreciated support of even fill, it seems that you might actually get placed in your secondary role more frequently than your primary. Is this a possibility? > > Are you also considering adding exceptions for the fill position? Personally, I feel comfortable on all the lanes, but I am an atrocious jungler. Is there anyway I could fill for any position that isn't jungle, or will I just have to choose 2 laning roles? > > Also, one final, more complicated question. > How, if at all, do you think the new process will alter the pick/ban phase? > > Thanks for your time! I'm loving his the changes are looking so far! We have done a few models internally of how each position will shake out. As you can imagine, players who choose Support or Jungle as their "Primary" position will most likely get them a majority of the time. If players choose a more popular primary position like Mid or ADC, it could be as low as 50% of the time; however, we have a few features in the matchmaker in place that will detect if you haven't gotten your Primary position for awhile, and bias the system to make sure you get it. We wanted to keep the system as simple and intuitive as possible for the first release, so there's no "exclusion" features yet for positions. We have a few sketches of how we'd do "Fill Except X" or "Choose 3 out of 5 Positions," but we'd like to keep it simple and see how the queue times and health is before we start adding more complex position selection schemes. In our internal labs where we brought pros and players in to test the new Distributed Ban system, we saw many players start banning out strong champions in their personal lane. This actually made lanes a bit more competitive because you could try crazier, off-meta champions and ban your counters.
Vanillite (EUW)
: Will we be able to choose ward skins and buy champion skins in the future during the new select?
> [{quoted}](name=yifei123,realm=EUW,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=000f,timestamp=2015-11-12T19:17:04.353+0000) > > Will we be able to choose ward skins and buy champion skins in the future during the new select? Yes! This is a feature we're still working on, but didn't finish by the first PBE release.
: By "priority champ" selection, do you mean the first/second picks can choose their champions before the ban phase begins?
> [{quoted}](name=DaRealWario,realm=NA,application-id=cIfEodbz,discussion-id=WV2VqZca,comment-id=0000,timestamp=2015-11-12T18:38:25.886+0000) > > By "priority champ" selection, do you mean the first/second picks can choose their champions before the ban phase begins? No, we just mean that the first and second slots get 1st and 2nd pick, which are traditionally powerful picks in draft. We have considered changing drafting more significantly, but didn't settle on something we liked for this pre-season.
  Rioter Comments
  Rioter Comments
: > [{quoted}](name=Lyte,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=004f0000,timestamp=2015-11-05T08:54:48.693+0000) > > Would love to, but generally don't like using data to try to convince players of something if we can help it. Wha-wha-what?! Isn't that saying "Believe us because we say so, not because of this proof we have sitting right here that you can't see."
> [{quoted}](name=Maestro80,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=004f00000001,timestamp=2015-11-05T14:12:54.596+0000) > > Wha-wha-what?! Isn't that saying "Believe us because we say so, not because of this proof we have sitting right here that you can't see." I've personally seen us post data before and have players say they don't believe it. So, I don't think data solves that issue.
Jaawn (NA)
: > [{quoted}](name=Lyte,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=004c0000,timestamp=2015-11-05T08:38:19.979+0000) > > We run random sample surveys across the entire global playerbase regularly, so we tend to get a nice representative response from all demographics--the hardcore, the average, the player that never visits forums, the player that visits Reddit 8x a day, etc. > > The truth is, in every survey we've done, the **majority** of players were interested in Dynamic Queue. That doesn't mean we just use that data point and force Dynamic Queue on everyone because we also saw that a non-trivial number of players were solo purists, and loved the idea of "true" Solo Queue and their experiences may get slightly worse because of Dynamic Queue. So, we assessed all the potential negatives that Dynamic Queue had, and implemented upgrades or new features that tried to address them. Did we succeed in making it overall a great experience for the vast majority of players? We don't know yet. We need to try, collect data, learn, and iterate. > > That's how design is. I mean I get that, I understand that this is why you're doing a progressive rollout to PBE and then Normals first. I'm just worried about selection bias and self-fulfilling prophecies. It is pretty easy to take a huge dataset and find data in it to support any preexisting hypotheses. And I get that this is your job, and you understand these things, but I have already noticed for a long time that there are issues with MMR and the matchmaker in lower tiers; I don't see how this won't make it worse. It feels like you're foreclosing on the idea of a truly accurate ranking system at all tiers. It's just "good enough." I guess I just don't agree that these tradeoffs are valid or "worth it." Even if your surveys did give you truly representative samples, and most players really do want this (and they specifically want it in solo queue)...does that mean it's the best option? You know, I bet most players would like free skins, or to be able to pay Riot for a fresh level 30 account. Maybe most players would even like to be able to buy "LP boosts" or be able to queue up for ranked with friends 3 tiers above them again. My point is, what people want and what is best for them or for the community are not always aligned. If most people want dynamic groups AND they also _specifically_ want them in solo queue...then I think this is one of the times where what people want is not what is best for them. By the way, if you already have a solution for the decreased precision in individual rank, I'd be very interested in hearing about it.
> [{quoted}](name=Jaawn,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=004c00000000,timestamp=2015-11-05T09:07:07.479+0000) > > I mean I get that, I understand that this is why you're doing a progressive rollout to PBE and then Normals first. I'm just worried about selection bias and self-fulfilling prophecies. It is pretty easy to take a huge dataset and find data in it to support any preexisting hypotheses. And I get that this is your job, and you understand these things, but I have already noticed for a long time that there are issues with MMR and the matchmaker in lower tiers; I don't see how this won't make it worse. It feels like you're foreclosing on the idea of a truly accurate ranking system at all tiers. It's just "good enough." > > I guess I just don't agree that these tradeoffs are valid or "worth it." Even if your surveys did give you truly representative samples, and most players really do want this (and they specifically want it in solo queue)...does that mean it's the best option? You know, I bet most players would like free skins, or to be able to pay Riot for a fresh level 30 account. Maybe most players would even like to be able to buy "LP boosts" or be able to queue up for ranked with friends 3 tiers above them again. My point is, what people want and what is best for them or for the community are not always aligned. If most people want dynamic groups AND they also _specifically_ want them in solo queue...then I think this is one of the times where what people want is not what is best for them. > > By the way, if you already have a solution for the decreased precision in individual rank, I'd be very interested in hearing about it. It's getting late--I need to get some rest so will skip some of your bigger discussions for now. At the risk of sounding offensive (which is NOT my intention with this reply), couldn't we use the same arguments against solo purists? What if Dynamic Queue is something solo purists say they don't want, but actually is better for them in the long run too? My point is, I agree that players sometimes say one thing, but mean another. Or, sometimes players really don't know what they *actually* want. But, I'd be wary assuming "the other population" is the one that doesn't know what they want, and *your* population knows exactly what you want and that you must be right. It's a game designer's job to try to figure out the reality behind the fog, and try to understand what is actually the best experience--but I'll be the first to say that it's hard, and sometimes we'll get it wrong. This might be one of those times, it might not be. But, we're willing to give it a go and pivot as necessary--even if it's mid-season.
Guiotine (NA)
: Any chance we could get some data after the new queue is released? I don't even mean as a way for players to "police" how well the system is working, I'm just genuinely interested in things like current toxicity levels in champ select vs. toxicity levels in champ select after release, current happiness/enjoyment of ranked vs. happiness/enjoyment of ranked after release etc. Mostly things to compare current values to values after the new queue is released. I realize the answer is almost definitely "no", but I would still like to see that kind of data :/
> [{quoted}](name=Guiotine,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=004f,timestamp=2015-11-05T08:47:05.813+0000) > > Any chance we could get some data after the new queue is released? I don't even mean as a way for players to "police" how well the system is working, I'm just genuinely interested in things like current toxicity levels in champ select vs. toxicity levels in champ select after release, current happiness/enjoyment of ranked vs. happiness/enjoyment of ranked after release etc. Mostly things to compare current values to values after the new queue is released. > > I realize the answer is almost definitely "no", but I would still like to see that kind of data :/ Would love to, but generally don't like using data to try to convince players of something if we can help it.
: I think the skill level and game knowledge required to be a top tier player has significantly diminished and will continue to do so with this champ select and queue change. I think back in season 2 you needed to have a firm understanding on how to play 5 roles to succeed in ranked. While now I can effectively need 1 lane worth of knowledge. This seems counter competitive at best. Do you have an opinion regarding this? Would you agree that the overall skill level has diminished in diamond players since season 2? And if so will these changes just facilitate the diminishing seen?
> [{quoted}](name=vampirekid13,realm=EUW,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=004e,timestamp=2015-11-05T08:34:37.009+0000) > > I think the skill level and game knowledge required to be a top tier player has significantly diminished and will continue to do so with this champ select and queue change. > > I think back in season 2 you needed to have a firm understanding on how to play 5 roles to succeed in ranked. While now I can effectively need 1 lane worth of knowledge. This seems counter competitive at best. Do you have an opinion regarding this? Would you agree that the overall skill level has diminished in diamond players since season 2? And if so will these changes just facilitate the diminishing seen? I don't believe skill level required to be a top tier player has diminished over time. In fact, it's increased over time. 1) When you look at Ranked data, most players are only proficient at 2 positions. 2.2, to be specific. There wasn't really a time period in League history where players were proficient at all 5 positions. In the new Champ Select, you still have to master at least 2 positions to rank up effectively. 2) As the game has matured, the skill level required to compete has only steadily increased. There are more websites, guides, blogs, deep analysis on game mechanics than ever before which is improving education of players on game mastery. There is more infrastructure in eSports (for highschool, colleges, and pro) than ever before, and it's greatly improving the 'average' skill level of top players. 4-5 years ago, there was no true understanding of 'optimal' ways to play the game, and you didn't have to understand little nuances to be in the Top 500 players in the world. Today, by about Level 15, players already are beginning to get a feel for the meta, what champions are effective at what positions, etc. A top player today would absolutely crush a top player from a few years ago.
Jaawn (NA)
: > [{quoted}](name=Lyte,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=,timestamp=2015-11-05T01:14:56.121+0000) > > ..But, right now the data suggests **A LOT** more players want Dynamic Queues than not ... What data? I see a lot of posts from people who don't want this. There is definitely a significant group of people who do want dynamic groups, so I'm not saying "nobody wants this." However, I don't think it is the majority of all players in solo queue (which, I'm sure you understand, is a different statistical population than "people who responded to some surveys" or "people who look at boards.na.leagueoflegends.com" or "twitter users" etc...) Also, it is fine if "**A LOT**" of players want this. You should definitely find a solution for them, but I don't think altering the solo/duo queue is the right call. Maybe this should be a modification to ranked 5s in tiers below Diamond. I listed a few ideas in a post I made (such as creating more robust team-growing tools and helping players find and create groups for ranked 5s). But this is going to turn solo queue into even more of a muddled mess in lower tiers. Even if you mitigate risks like premade 4s bullying solos, without HUGE advances in your ranking and matchmaking algorithms, you might as well combine Bronze, Silver, and Gold into one "Hell" tier. It can be the mirror of Master tier, where every night the top 25% of players get moved up to Plat, and everyone else continues the battle royale. Maybe this is a slight exaggeration, but it is seriously going to move further in that direction without a **robust** individual skill/rank tracking system that goes far beyond win/loss. I feel like I could explain a lot of reasons why this is going to create problems for MMR/matchmaking/general rank quality or how it is going to be a negative experience for introverts and less verbal players...but you all seem to have already made up your mind :-\ Let me just simply say: I am a customer. I have been for almost 6 years now. I do not want this.
> [{quoted}](name=Jaawn,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=004c,timestamp=2015-11-05T08:15:24.239+0000) > > What data? I see a lot of posts from people who don't want this. There is definitely a significant group of people who do want dynamic groups, so I'm not saying "nobody wants this." However, I don't think it is the majority of all players in solo queue (which, I'm sure you understand, is a different statistical population than "people who responded to some surveys" or "people who look at boards.na.leagueoflegends.com" or "twitter users" etc...) > > Also, it is fine if "**A LOT**" of players want this. You should definitely find a solution for them, but I don't think altering the solo/duo queue is the right call. Maybe this should be a modification to ranked 5s in tiers below Diamond. I listed a few ideas in a post I made (such as creating more robust team-growing tools and helping players find and create groups for ranked 5s). But this is going to turn solo queue into even more of a muddled mess in lower tiers. > > Even if you mitigate risks like premade 4s bullying solos, without HUGE advances in your ranking and matchmaking algorithms, you might as well combine Bronze, Silver, and Gold into one "Hell" tier. It can be the mirror of Master tier, where every night the top 25% of players get moved up to Plat, and everyone else continues the battle royale. Maybe this is a slight exaggeration, but it is seriously going to move further in that direction without a **robust** individual skill/rank tracking system that goes far beyond win/loss. > > I feel like I could explain a lot of reasons why this is going to create problems for MMR/matchmaking/general rank quality or how it is going to be a negative experience for introverts and less verbal players...but you all seem to have already made up your mind :-\ > > Let me just simply say: I am a customer. I have been for almost 6 years now. I do not want this. We run random sample surveys across the entire global playerbase regularly, so we tend to get a nice representative response from all demographics--the hardcore, the average, the player that never visits forums, the player that visits Reddit 8x a day, etc. The truth is, in every survey we've done, the **majority** of players were interested in Dynamic Queue. That doesn't mean we just use that data point and force Dynamic Queue on everyone because we also saw that a non-trivial number of players were solo purists, and loved the idea of "true" Solo Queue and their experiences may get slightly worse because of Dynamic Queue. So, we assessed all the potential negatives that Dynamic Queue had, and implemented upgrades or new features that tried to address them. Did we succeed in making it overall a great experience for the vast majority of players? We don't know yet. We need to try, collect data, learn, and iterate. That's how design is.
Thales (NA)
: > [{quoted}](name=Lyte,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=001b00000000000000020001000400010001000100000000,timestamp=2015-11-05T05:58:52.592+0000) > > You're right, but I didn't want to dig deep into the math weeds again :P > > The key is in the numbers. To go super specific, if we expect a Mid-player to "carry" the game in X% of their matches, and the player only "carries" and impacts their game in X-10% of their matches, then something is off. The key is always what the actual %s are compared to our expected %s. So, on a 4:1 team, a common complaint is that the solo is unable to influence the game. This is a *prediction*, and we can see in the data whether this is true or not. If it's true, and solo players with a Premade of 4 can no longer effectively influence the outcome of a match, we do agree that's a problem and we'll try to fix it. Heh, fair enough. After this finally goes live, I'd be really interested to learn how similar the premade behaviors are in it as opposed to normals. My intuitive guesses are that the solo players in a 4-1 team will prove to have more of an influence on how the game goes in ranked than in normals, and that teams of 5 will have a larger advantage when paired against teams that aren't fully premade in ranked than they do in normals, but I'd love to learn how things shake out once once you actually have some numbers. Actually, since draft normals and ranked currently use the same format, it might be interesting to see if patterns and relationships currently existing between them carry over when they're replaced with this new system.
> [{quoted}](name=Thales,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=001b000000000000000200010004000100010001000000000000,timestamp=2015-11-05T06:08:48.967+0000) > > Heh, fair enough. > > After this finally goes live, I'd be really interested to learn how similar the premade behaviors are in it as opposed to normals. My intuitive guesses are that the solo players in a 4-1 team will prove to have more of an influence on how the game goes in ranked than in normals, and that teams of 5 will have a larger advantage when paired against teams that aren't fully premade in ranked than they do in normals, but I'd love to learn how things shake out once once you actually have some numbers. > > Actually, since draft normals and ranked currently use the same format, it might be interesting to see if patterns and relationships currently existing between them carry over when they're replaced with this new system. This is exactly why we'd like to work in Normal Draft Queues first, because they are a close model of how a Dynamic Queue *might* work. We're also testing "Ranked Queues" in NA and TR, to get an even closer model of how things might shake out. We'll know more soon, we just need players to give us a chance to try things before jumping up in arms and condemning it. We aren't crazy, and we know this is risky--but, we want to try.
Thales (NA)
: > [{quoted}](name=Lyte,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=001b000000000000000200010004000100010001,timestamp=2015-11-05T05:22:56.186+0000) > > We've dug into examples like this many times before, and it's generally because the Normal MMRs of the players don't reflect their Ranked MMRs. This is a known problem that we have some fixes for called "Splashing." Basically, as your Ranked MMR increases, we "splash" some points into the other queue MMRs to make sure they are all tied a bit closer. But, in practice this didn't work too well because a lot of higher tier players mess around in Normals or only want to play casual in Normals with friends, so they DIDN'T want their Ranked MMR to affect other queues. > > Long story short, it does look bad, but actually if you look at the matches holistically, you still have very close to a 50% win rate in those matches. I'm not sure if a 50% win rate really supports your conclusion. If every game were decided by which team gets the master tier Yi, then people would have close to a 50% win rate, but would have very little influence over whether or not they win their games. When I play, I want to feel like how I play will affect the game, rather than have it decided at matchmaking. A coin toss is fair, but it's not fun. Then again, if a player is a guaranteed win in normals, their normal MMR should hopefully rapidly rise to the point where they don't run into lower tier players. And I do know that many very good players use normals to mess around: a few months ago, I faced a pro playing Sated Devourer jungle Lucian (I won! Barely...). Your conclusion may be right, but I'm not sold on 50% win rate indicating much.
> [{quoted}](name=Thales,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=001b0000000000000002000100040001000100010000,timestamp=2015-11-05T05:51:05.116+0000) > > I'm not sure if a 50% win rate really supports your conclusion. If every game were decided by which team gets the master tier Yi, then people would have close to a 50% win rate, but would have very little influence over whether or not they win their games. When I play, I want to feel like how I play will affect the game, rather than have it decided at matchmaking. A coin toss is fair, but it's not fun. > > Then again, if a player is a guaranteed win in normals, their normal MMR should hopefully rapidly rise to the point where they don't run into lower tier players. And I do know that many very good players use normals to mess around: a few months ago, I faced a pro playing Sated Devourer jungle Lucian (I won! Barely...). Your conclusion may be right, but I'm not sold on 50% win rate indicating much. You're right, but I didn't want to dig deep into the math weeds again :P The key is in the numbers. To go super specific, if we expect a Mid-player to "carry" the game in X% of their matches, and the player only "carries" and impacts their game in X-10% of their matches, then something is off. The key is always what the actual %s are compared to our expected %s. So, on a 4:1 team, a common complaint is that the solo is unable to influence the game. This is a *prediction*, and we can see in the data whether this is true or not. If it's true, and solo players with a Premade of 4 can no longer effectively influence the outcome of a match, we do agree that's a problem and we'll try to fix it.
Healin (NA)
: > [{quoted}](name=Lyte,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=001b0000000000000002000100040001,timestamp=2015-11-05T04:26:57.965+0000) > > When players make statements like "Skill Ranking" means nothing, it might not actually reflect reality. > > snip My post has been eaten 3 times now so sadly this will be the condensed version of it. I feel you got distracted by a tangent of the math behind Higher leveled players whomping on lowbies. You don't seem to even acknowledge if it's okay for that even to happen. Your current matchmaking seems to fail on regular basis. After keeping track for awhile I noticed some pretty awful matches where it seemed that instead of trying to give me a 50/50 match based on my own skill, it instead used me as a lead weight to bring down the average level of much better players. Take this egregious example. http://i.imgur.com/IbZs1P5.png First of all, the absolute insane spread of skill levels is what shocks me here. We have Unranked, Silvers, Platinums and Diamonds all in the same game. What's even worse? Take a look at the enemy Yi. As a silver 4 player, I should never have to go up against a Master Tier Yi. It just simply shouldn't happen under any queue. It doesn't even matter how he did that specific game. Something in Matchmaking should have been sending up red flags saying "Whoa there, something doesn't seem right about this" I'm fine with losing over and over in a fair match. But being used as an albatross to punish other players is something else entirely. And unless you solve this problem of the solo players being used as a handicap for the benefit of quicker queue times for people who duo/tri/quad queue then you're not going to stop seeing people complain about it.
> [{quoted}](name=Healin,realm=NA,application-id=mNBeEEkI,discussion-id=JNwgrzih,comment-id=001b00000000000000020001000400010001,timestamp=2015-11-05T04:59:10.893+0000) > > My post has been eaten 3 times now so sadly this will be the condensed version of it. > > I feel you got distracted by a tangent of the math behind Higher leveled players whomping on lowbies. > > You don't seem to even acknowledge if it's okay for that even to happen. > > Your current matchmaking seems to fail on regular basis. After keeping track for awhile I noticed some pretty awful matches where it seemed that instead of trying to give me a 50/50 match based on my own skill, it instead used me as a lead weight to bring down the average level of much better players. > > Take this egregious example. > > http://i.imgur.com/IbZs1P5.png > > First of all, the absolute insane spread of skill levels is what shocks me here. We have Unranked, Silvers, Platinums and Diamonds all in the same game. > > What's even worse? Take a look at the enemy Yi. > > As a silver 4 player, I should never have to go up against a Master Tier Yi. It just simply shouldn't happen under any queue. It doesn't even matter how he did that specific game. Something in Matchmaking should have been sending up red flags saying "Whoa there, something doesn't seem right about this" > > I'm fine with losing over and over in a fair match. But being used as an albatross to punish other players is something else entirely. And unless you solve this problem of the solo players being used as a handicap for the benefit of quicker queue times for people who duo/tri/quad queue then you're not going to stop seeing people complain about it. We've dug into examples like this many times before, and it's generally because the Normal MMRs of the players don't reflect their Ranked MMRs. This is a known problem that we have some fixes for called "Splashing." Basically, as your Ranked MMR increases, we "splash" some points into the other queue MMRs to make sure they are all tied a bit closer. But, in practice this didn't work too well because a lot of higher tier players mess around in Normals or only want to play casual in Normals with friends, so they DIDN'T want their Ranked MMR to affect other queues. Long story short, it does look bad, but actually if you look at the matches holistically, you still have very close to a 50% win rate in those matches.
Show more

Lyte

Level 30 (NA)
Lifetime Upvotes
Create a Discussion