Rethinking Ranked Fives and Tuning Dynamic Queue

Two weeks ago, we confirmed we're not bringing back solo queue and promised to do a better job of keeping you updated about our ongoing efforts to improve ranked.

So let's talk about what we're doing to address three specific issues:

• Recognition for skilled solo play in dynamic queue
• Opportunities for ranked fives
• Matchmaking problems including queue times for high MMR players and improved odds for everyone to get their primary role

INDIVIDUAL ACHIEVEMENT AND RECOGNITION IN DYNAMIC QUEUE

Individual achievement isn't given the same recognition in Dynamic Queue, and this has left some solo-only players feeling disenfranchised. We're shipping a feature in 6.13 that ought to help: ranked emblems.

The goal is to give you another tool (alongside your rank and champ masteries) to show what sort of player you are. We'll track your win history and premade size, and once you reach 25 wins in Dynamic Queue you'll earn either a Solo Emblem, a Dynamic Emblem, or a Team Emblem. The emblem you earn is based on the way you've queued up for your previous 25 wins, and you can only have one emblem at a time. Emblems have no impact on matchmaking, but they'll show up on the ladder and will be accessible to third-party tracking sites through our API.

So how do you know which emblem you'll get after 25 wins?

• Solo Preference Emblem: Queue alone for most of your games
• Dynamic Preference Emblem: Queue with varying team sizes for most of your games
• Team Preference Emblem: Queue with full teams for most of your games.

The ranked emblems system isn't retroactive, so it'll begin awarding emblems based on the first 25 wins you get after it launches.

With solo queue gone, there needs to be better recognition of individual achievement. We're committed to solving this problem, and we think this implementation of ranked emblems probably won't completely solve this issue for the players who are least happy with dynamic queue. We're seeing emblems as one step towards a better system, not as the end-all-be-all solution to the problem. We want to see how this is received before we move on to other options (and we are looking at other options).

SCHEDULED RANKED 5v5 TEAMS

We disabled ranked teams at the start of the season because we were unsure how match quality would be affected by the addition of dynamic queue. Now that we've limited high MMR players (Diamond 5+) to solo, duo, or trio queues, those players have no option to play ranked as a group of five. Even among the much larger population of players who fall below Diamond 5 (about 99% of all players), we've heard that people miss the ranked 5v5 queue. They miss the unique challenge of managing a consistent ranked team without the constraints of tier restrictions. Queue population issues prevent us from bringing back a permanent ranked fives queue, because we can’t guarantee reasonable queue times in off-peak hours. Instead of giving up on ranked fives, we want to try a potential solution: scheduled ranked 5v5 teams.

Other than the fact that this queue will not be available 24/7, it'll function exactly like the old ranked 5v5 teams queue did. We're still evaluating exactly what time and days for each region make sense, so look for more info on this in the coming weeks.

QUEUE TIMES AND POSITION SELECT

Queue times have been far too long for high MMR players. As part of our ongoing efforts to solve this, we introduced an autofill mode that triggers when we expect position select to lead to unacceptably high queue times. Our initial data shows that high MMR queue times have improved dramatically, and most players in Master and above are only ever impacted by autofill once every ten games. Here are more specific stats:

You'll notice that queue times for everyone under Diamond tier have actually gone up slightly. This has nothing to do with our tuning of the autofill feature. Instead, this is a trade-off we're making to increase the likelihood you'll receive your primary role in champ select. Regardless of your ranking, you should have a better chance to get your primary position.

These are great gains, but what about the criticisms against autofill? The autofill feature undermines one of the main promises of new champ select—that you'll get the role you want. This is undeniably true. Forcing players into autofill means they'll sometimes get stuck in their weakest role, and this makes climbing into the highest ranks as a one-trick or mid-only player much harder.

We're okay with this right now for two reasons: 1) It's necessary to get high MMR queue times down and improve matchmaking quality, and 2) Players dealt with an even more drastic form of randomized role selection for years in the previous champ select process. The high queue times and matchmaking inequalities we've been dealing with resulted from our switch to new champ select with its position-select features.

But we aren't willing to give up the gains that new champ select brings us—it's still awarding high MMR players their desired role at a much higher rate than what players were used to with last season's ranked draft pick. Instead, we'll continue making adjustments and improvements that minimize the irritation of not getting your preferred role.

The shift to dynamic queue is about reinforcing League as a team game, so organized team play needs to be much, much easier than it is now. We need to offer the right platforms, tools, and team-finding capabilities to stand behind League as a team sport. We're not there yet, but we're determined to do the work get there.

This is an iterative process. We'll keep communicating specific details about our plans for improving ranked, and we'll be honest about what's working and what isn't. Ultimately, we're aiming to give every type of player (including solo only players, high MMR folks, and people who are currently happy with dynamic queue) the best League experience possible.