Boards›Gameplay

Suggestion to make Dynamic Queue Fair!

Dr Poro··42 votes·1,946 views

Current situation

  • Solo Queue is unlikely to come back
  • There is quite a dissatisfaction with Dynamic queue. Two issues I recognize are:
  1. Amount of premades per team isn't balanced out (Riot claims to be working on this currently)
  2. Ranks lose their value when people can easily get carried by their (smurf) friends, which is what I will focus on in this thread.

Why is this a problem? Getting carried by premade friends throughout numerous games into a division that you don't deserve, is basically a cheat on the ranked system. Ranks are supposed to represent your skill as a player (in the broad sentence: mechanics, knowledge, awareness, teamplay, communication, adaptation, etc.) NOT your friends' skill level or the size of your wallet.

Can you prevent this? It seems nigh impossible to prevent smurfs or powerlevelling without hurting other players with it, but there are ways to make it a bit harder for them to do so. What I suggest is a way to make the luggage (the one getting carried) gain less MMR while the carry gets more MMR from victories.

The Grades system, while not perfect in determining individual skills (communication, selfless plays for the team, ...), it does cover a broad area of it and is a pretty good predictor of who performed well and who didn't (way better than KDA). I suggest that WITHIN a premade team, in victory players gain more or less MMR than usual based on their Grades relative to their premade friends.

MORE CONCRETE: Each premade team will get a pool of MMR which is equal to what is currently the sum of the MMR each member would get. This pool will then be distributed among the party members based on the performance (Grades) relative to the expected performance (higher MMR players are expected to perform better than their lower MMR friends).

what does this mean? When you win a game as a premade team, the members of the party will gain MMR based on their contribution in victory. Members that performed poorly throughout the game will be rewarded less than members that 'carried' the game. A part of the MMR that would currently go to the low performing player will now go to the good player.

Example _Three man premade: Gold I, Silver IV and Silver V. Gold I gets S-, Silver IV gets C-, Silver V gets S-. The MMR gains would be Silver V > Gold I > Silver IV with Silver V far above average (I would say around x1.5 gain, but numbers are up to specialists), a slight increase in gain for Gold I and a vast decrease for the Silver IV's MMR gain (e.g. x0.3). _ This is an extreme situation where it's obvious that the Silver IV got carried in this particular game. Is it because he's 'bad'? Who knows, but the MMR differences will have an increasing impact in the long run if this continues to happen.

Why MMR and not LP? LP is a short term factor. LP changes are based on your MMR, which is the underlining red wire of your progression. Getting carried by friends is not a short term problem. It's okay to win a game or two because your friends had an amazing game. It's not okay to gain 2 tiers because your friends are amazing players on smurf accounts. If you play with a friend of equal skill and equal MMR, then you will both have good and bad games and everything evens out in the end. A rollercoaster of LP differences would be frustrating. If a player teams with a smurf, the smurf will consistently play better and reaches his 'true MMR' quicker than currently while the luggage didn't climb as far. While you can ask a friend to carry your promotion, this is only a short term and short distance advantage that you cannot stretch infinitely.

Why only in victories? Two reasons why I added this detail.

  1. The main issue is when players team up with players that they know will be likely to win. They will not lose as many games as most people during this power level. Affecting defeats would be a very aspecific way to deal with this as it affects other players more.
  2. Currently I feel that the Grades system isn't in place yet when it comes to grading the losing team. Players that try to protect their base/team with their lifes get punished for their high death score.

What does this change for solo queue players? Nothing. (Except hopefully relief some frustration and a fairer chance at climbing because of smurfs reaching their rank faster)

Why not make one MMR pool for the entire team? There is no reason to affect solo queue players with this. Winrate for solo queue players is based on their own performance and RNG. The performance of their team and the opponent team is one big RNG factor that will even itself out in the long run. This is not true for premades. You can keep playing with the same friends and their skill/performance (or lack of it) will never get compensated. The system has to do this in another way (I suggested one way to do it.)

Pitfalls This is not a feature to take lightly. It's a very fragile system that falls by the slightest bias.

  1. The grade system has to accurately represent the performance. I'm aware that there are currently champions/playstyles where this isn't the case and future issues can arise which requires proper monitoring of this system.
  2. The MMR modifier has to be fair. It may not be too harsh and not too light for maximum utility.
  3. Expected performance between players of different MMR has to be carefully considered and isn't an easy task.
  4. This system may not be abusable in the other direction (searching for a player that intentionally gets a bad score in games where victory is guaranteed.) This behaviour should be recognized and nullified.

TL;DR? The Bold paragraph starting with 'more concrete' is the essence. For more context, I suggest reading the whole text.

Let me know what you think. Constructive critism and open discussions are very much appreciated.

39 Comments

D rok9 votes

I appreciate the effort you've given into a post like this, but I really don't think this is possible, for a few of reasons.

A: The Mastery system is really not a great factor in determining how well a person plays, especially at low levels of play, and especially this season, when counters tend to be more prevalent, and dynamic queue can easily make 1 lane more favoured than another (Jungle and top queued together will favour ganks against a top without a jungler queued with him, for example). A fiora or tryndamere who is 2-10-2, can still provide a massive split pushing advantage that can win you the game (especially in low elo). A clutch shen ult can save a game, regardless of how well he did in lane. Late game champions tend to do poorly in the early game, and their MMR shouldn't be affected by that. Lots of reasons mastery just isn't the right system for this.

B: The mmr system is also just... so hidden. It would either make so little difference that you wouldn't recognize it, or conversely, provide a disincentive to play with friends who aren't straight up better than you.

C: I think, at least in your example, you may be overestimating the difference in expected play between a Gold 1 and a Gold 5. It's pretty minimal, honestly.

I've seen a lot of suggestions for changing mmr between solo queuers and team queuers, and I just don't think it will help. The solo queuer is always going to be in the same system as those queued on a team, which is always going to provide the problems people are talking about, beyond just climbing.

JVP151 votes

A: The mastery system isn't great, but maybe it can be improved? So, say in lane, you do poorly. But then later in the game, you just unload a ton of damage. Your CS might be a tad lower, but it is made up for by the fact that you just did a crapton of damage to towers, dragon and baron, and champions. For supports (and Shen, and Rammus, and Malphite) you could also factor in the amount of damage they tanked vs the amount of times they died, the amount of times they healed, the amount of shield give, and the amount of CC they gave.

B: The MMR system is pretty hidden, yeah. But, the question is, are you trying to play against better people or simply trying to get a better rank (and therefore a cooler border) by playing with your smurf friends. I am not talking about you you, but the you as in the people who are playing with their smurf friend to boost their rank. So, MMR is hidden. However, LP isn't, and it seems like LP determines what rank you are (if I am wrong, please correct me) so maybe the amount of LP you get from a win can be determined by what he said above.

C: From what I can tell, the friends of the Gold 5 where smurfs that went down to Gold 1. I believe the smurfs are actually something like Platinum or Diamond. It is just a guess, but I thought the thread was trying to say that the players from Gold 1 where smurfs from a higher tier.

Akeydel1 votes

Just a counterpoint to B (which i believe to be the heart of the argument) There are so many reasons why you or your teammates can get different LP gains/losses, having this hidden system wouldn't have a noticeable impact, (it may already exist in some form!), unless you make it so significant that people go back to the season 3 complaint about LP clamping, which Riot made a point to remove.

RevokeFaith7 votes

Wow, this is the best idea I've seen so far. I don't think it's easy to exploit it because if you intentionally try to get a lower score as a smurf, you'll lose the game more often which is something you can't afford. The system has to be made so that doing things like these are never profitable, though. I like how it keeps the game the same for soloQ players while allowing premades to exist but without the 'get carried' aspect. It's probably still possible to get carried, but way harder. Maybe even harder than before (with duo queue).

J E B A ł T E D4 votes

I Kinda hope Repertoir can chime in on this tbh. As of now, even though games themselves can be fun, the fact you can premade and win it all without any drawbacks kinda makes solo climb pointless to me.

Dr Poro1 votes

Yes, I'm also very interested in Repertoir's point of view. I'm sure he can see some holes in this version of the idea, but I hope he likes it and that Riot can do something with it.

MewTas3 votes

Good to see an actual constructive thread with a solution, rather than ten more rage posts about quitting the game. I would have no problem with dynamic queue if these 2 issues were solved.

Fenikkusu11 votes

The sad part is that unless it gets more upvotes Riot won't take notice. They never seem to see the constructive threads, or really any for that matter. So lets upvote the hell out of this post to make sure it will hopefully get seen.

Alüe3 votes

Don't mind what you suggested, still want solo q though on the off chance the pre-made on my team sucks compared to theirs.

Dr Poro2 votes

I understand. I've played since season 2 and always advocated to REMOVE duo queue from ranked. At this point, however, it seems that Riot's vision on the ladder is very different from mine and I don't see any use in hoping for them to change their minds. Instead, I'm trying to suggest a way to make their version more fair.

Premades is a big random factor, which is a lot harder for the system to predict their true rankings than soloQ players causing more variety in your teammates' and opponents' skill. I don't like this, but Riot doesn't seem to mind. In the long run, this RNG factor will even out just like soloQ teammates, but it takes a bit longer. As someone who plays less than 100 ranked games per season, it doesn't quite even out for me.

Alüe1 votes

i'm not entirely sure if they view Hots a threat, however since that's how blizzard adopted there ranked system (works out very well for them) i think riot feels compelled to do the same. However I think it's important for riot to understand it works well for Hots because it is so much more team orientated than league is.

Dr Poro1 votes

I'm not familiar with HotS but I think League of Legends is great because you can play it in different ways. You can focus on personal performance and winning the game by winning your lane first and extending your lead to the rest of your map, or by coordinated teamplay and win by strategic manouvres (the way it's usually played in tournaments). The majority of the players don't have the skill nor synergy to win purely based on these manouvres. In reality, I suppose you need a bit of both. With just mechanical dominance, you can climb, but you'll quickly find a roof if you don't go for objectives when you should.

sorry, I've gone quite off-topic here. I was just saying that League is a team game as well, but doesn't strictly have to be played that way. There are different ways to climb. There's probably one ultimate way to get to the top, but if your aim is gold or platinum, there's definitely different paths and it doesn't (shouldn't) require you to go premade.

ff4sk2 votes

Good idea. Not that riot or the people playing in premades will care though, as long as it benefits them fuck the solo players

Dr Poro1 votes

As a company, I'm sure that they want their players to be happy. Happy players spend more money and invite more of their friends to their game.

Always Be Polite1 votes

Very good thread. I wonder if Riot already thought about this.

Rickon Stark1 votes

So you're basically saying if you have a bad game you lose MMR or don't get as much as someone else. I didn't want this game to turn into CoD, but okay...

Dr Poro1 votes

You'll never lose MMR, the minimum MMR you get should still be significantly above 0, perhaps half of what you'd expect to get if you do poorly and your premades do well. Note that this also means that your premades will get more MMR. NEVER will you lose MMR to random players. SoloQ players receive their own pool and because they're the only player in that pool, they always get the same MMR as they would currently. If you duoQ with your friend and he snowballs his lane and wins you every teamfight while you never attended any teamfight and end up 0/11/2 with barely any cc, no turret damage dealt and 1 ward placed all game then yes, your friend should get a lot more MMR than you if you were both at the same MMR.

Why? Because it's obvious that this game wasn't won by you. It was won by your friend. He should get rewarded for it. Why not implement this for all players (soloQ as well)? Because teammates are a random factor. They sometimes carry, sometimes feed. You can't choose them, therefor, you don't need to compensate for it. You can choose your premades. You can choose to play with people that carry you consistently. That's why the Grading reward system is appropriate here as compensation for this non-random factor. Why? Because ultimately, your ranking should depend on one thing: Skill. In the wide sentence (this includes mechanical skill, awareness, communication, adaptation, etc.) Random factors will even out. Premades won't.

This system won't prevent premade carrying, but it will make it significantly harder.

  1. The luggage doesn't climb as quickly
  2. The smurf will climb quicker, which means that if he continues to do this, he'll reach his true MMR quicker.

Note that when you just play with friends of equal skill, you'll all have good and bad games sometimes and it'll even out. This will only have a huge impact when people are performing consistently above or below their MMR expectations while premading.

The Bíg Ticket1 votes

Except score or grades don't matter if you wanna determine who contributed how much to a victory.

Dr Poro1 votes

Grades are supposed to express how well you did in that game. Doing well = contributing to victory. In my opinion, the current grade system is pretty consistent, especially in victory. In defeat it's often poorly balanced because people that try to defend objectives and die a lot will get a lower score because of this. Also, it's very hard to get a high score with an all-in support if your team does badly. Perhaps grades shouldn't be used in defeat, only in victory, which might be enough to disencourage the smurf-carrying abuse.

Keep in mind that this only matters within the premade group.

The Bíg Ticket2 votes

Nah...the grade system doesn't really show how much a person contributed to a victory. Take the Lux mid for example who just sits mid and waveclears and in teamfights she stands back and does damage from range. She ends up with a 7 0 15 score and gets an S or S+. Meanwhile your Olaf or whatever jungler counterjungles the enemy Lee, gets all objectives for your team, makes good calls, pushes the right lanes at the right time and in teamfights he always dives the enemy APC and ADC which leads to them being completely out of the fight so Olaf's team has nothing to worry about except dealing damage. However he died a couple times by dying that and now his final score is 5 8 10. He will now get a worse grade than the Lux even though her arguably contributed atleast the same amount, if not alot more.

Grades are not a good way to determine how much LP each person should get.

Odinakos2 votes

Expanding on that. It is a lot easier to get S rank on safe range supports rather than tank supports. I say this because resently i was trying to get S rank for boxes.

In order to get S rank with Allistar i had to go through lot of A rank games when i performed exceptionaly(i will discount the lower ranks since i didnt deserve it). Being in frontline causes some deaths naturally but also had on average 2 deaths per game when i sacrifised myself for the team (like salvaging a lost teamfight, peeling for allies that were caught). And there was the decision of mastery rank vs the game. While on Nami (which wasnt even on my top 5 supports) it took me a lot fewer games to get that S rank.

So from my prospective. I can easily spam Janna(my best support) all day long for easy A and S ranks. By no means i say i am a Janna god, only that is very easy to get high ranks even if another support would be better and provide more influence in my games. That would mean in certain games i would make an inferior choise in order for me to have better LP gain/lose.

The Bíg Ticket1 votes

Yup, that is another good example.

Aazzlano1 votes

What... the hell? Yes, the mastery system is not indicative of how much a person contributed to victory, but your example is ridiculous.

You are really delusional if you think that a 5/8/10 Olaf did even remotely as well as a 7/0/15 Lux. Lux obviously did her job by waveclearing and dealing damage from far back, and played safe enough while doing that, securing a lot of kills, and had a high kill contribution overall. Meanwhile Olaf over here has barely as much total kill contribution as Lux does as assists, and having over half of that number in deaths too. He may have done his job in teamfights in the end but he didn't contribute as much and she did, AND did so unsafely. On top of Olaf having a bad score (not indicative of not contributing however) you said that Lux also wavecleared mid while Olaf killed monsters in the Jungle. That means Lux put wave pressure, took pressure off her towers, in addition to playing safe and getting kills, picks, and peeling the rest of the backline, on top of shielding her team as well. Meanwhile Olaf counterjungled the enemy Lee who.... doesn't care about counterjungling and just ganks anyway? Sure he may have got all the objectives in your fantasy scenario but Lux was the one clearing waves and preventing her laner from roaming, opening up the possibility for Olaf to even take objectives.

Champion mastery is rated on how well you did versus other players of that champion in that role. So that Olaf got a shit champion mastery because he did shit. Other Olafs do much better than he did. That's how it works. If he spent that much time counterjungling and was unable to be strong enough to not die 8 times then that is a problem with the player, not the rating system. Maybe if the enemy carries are so strong he should have spent more time ganking them and keeping them down, since his jungling obviously didn't make him strong enough, or the items he bought or way he plays is just bad. Obviously the system isn't perfect but you thinking that he would deserve an S for that is laughable. I wouldn't want that person on my team at all much less rate them an S for that gameplay. S ranks are for exceptional performances, he did not perform exceptionally.

The Bíg Ticket1 votes

Lux wavecleared mid while Olaf pressured all lanes plus the enemy jungler...if you had read what I wrote you could have saved that wall of text.

Aazzlano1 votes

Oh so he counterjungled, pressured all lanes, took all objectives, and still ended 5/8/10 ?

Your scenario is completely made up and unrealistic, sorry

The Bíg Ticket1 votes

Lol of course its made up...did you even read my post? You can do all of that and still get a "bad" score. Which is the reason that KDA and grade means nothing.

Aazzlano1 votes

You can make up any situation to fit your needs. You can say someone was the most amazing player, did everything and carried 100%, but was 0/19/0, and thus got a bad grade when they deserved S+.

You are making a fantasy situation which perfectly fits your needs where this guy was the greatest player on earth, doing absolutely EVERYTHING in the game yet getting a bad score because it fits your agenda, which is that it won't give him a good champion mastery.

The Bíg Ticket1 votes

Lol...its an example situation...calm down homie. You can do well but have a bad score...thats what I am saying. If I am dying to get turrets, zone off enemy carries and just make the right calls in general, I can easily end up with a bad score (and grade) even though I contributed a huge amount to our win.

Aazzlano1 votes

Yeah because you did badly.

If you did exceptionally well and deserving of an S... you would do your job without feeding.

Mandang01 votes

I think it's an interesting idea, but I can already think up a way to abuse it - instead of grabbing friends of a higher MMR to boost you, now you grab friends of a lower MMR to boost yourself. Ensure you get to lane/play against lower MMR opponents to pad your stats to compensate for potentially lower MMR gains for the win based on your being the highest MMR player on your team.

Either way, I think the fundamental problem is that dynamic queue undermines the quality of a matchmaking system that's already strained by trying to apply the outcome of a 5v5 match to the individual MMRs of each player in that match. Once you introduce dynamic queue, now you need fudge factors for teams of different-sized premades, teams with different degrees of MMR split within their premades, you're punishing players who don't set up their own voice chat (which is a problem because Riot doesn't provide in-game voice chat). And, in the case that you create a game with 2 lanes having a large MMR mismatch (each premade's best player against the other premade's worst), you're punishing the team that doesn't put their best player in the most important lane...

It all boils down to the fact that matching a bunch of individual players for a 5v5 match is a lot easier and more accurate than trying to match partial groups in dynamic queue.

AwesomeChad1 votes

YES, i agree with this model! Make a complex formula that uses individual performance as one of the factors in determining how much mmr a premade player gains in each game. It shouldn't be the same in the first place. Because riot wants to promote playing with your friends in ranked , they're created an uncontrollable unbalance in player skill due to players not being in their own leagues.

Zantumall1 votes

Wait, the "news that solo queue isn't coming"? There hasn't been any official confirmation that we're not getting solo queue to my knowledge. I'm pretty sure that's just forum circlejerking.

Mickey Donnerudo2 votes

A Korean Rioter has confirmed that solo queue has been "put on hold indefinitely".

Lauchmelder1 votes

Which doesn't mean it has been cancelled until they particularly say: "SoloQ has been cancelled completely."

Mickey Donnerudo1 votes

What's stated above is essentially business-speak for "cancelled". You never want to actually say "cancelled" because if you do end up wanting to release it in the future (e.g. the community gets mad as fuck) then you can just say "Sorry for the delay! Solo Queue is finally here!" as if you were planning it all along.

Lauchmelder1 votes

Which is essentially what I was saying. SoloQ isn't cancelled until it is cancelled. The customer interprets this as "cancelled" while it actually is put on ice for a certain time.

Mickey Donnerudo1 votes

If that's how you want to think of it.

They promised us solo queue though. And they are not planning on delivering.

It's cancelled.

T RexHasTinyArms1 votes

While I agree there are issues, I think part of the idea behind groups being added to ranked in the first place is to reward chemistry and teamwork. Leadership doesn't show up in the mastery ranking, and if we are assuming that a player who is a good shot caller or communicator is not earning their DQ rank through that skill w/ their party, then we really might as well scrap DQ. If all it is about is in game performance, then take the new champ select and go w/ solo only play which rewards mechanics more.

Unfortunately I think riot wants to give low elo scrubs like me the opportunity to play out the "I'll be like Hai and make my team better even if I don't dominate my role" fantasy and so DQ is thrown at us. At least that's the reasoning I've seen behind why it is a thing (otherwise it's just elo boosting). This solution doesn't capture that so prob an issue.

Dr Poro1 votes

I have rewritten the entire thread to make it more comprehensible. Wanted to change the title as well, but I'd have to make a new thread for that. I'm happy with the attention this thread got so far so I'll leave it as it is for now.

More from Gameplay

3 Refunds per season, upvote if you want 3 refunds per season.
Gameplay764 comments
7,167
votes
3 Refunds per season, upvote if you want 3 refunds per season.
Gameplay156 comments
3,510
votes
UPVOTE FOR URF
Gameplay606 comments
2,440
votes
Yes plz.
Gameplay456 comments
1,721
votes
3 Refunds per season, up vote if you want 3 refunds per season.
Gameplay114 comments
1,716
votes
When is ANYONE at riot going to address the awful new player experience?
Gameplay634 comments
1,383
votes

More by Dr Poro

PETITION Make 'First Win of the Day' Daily
Gameplay142 comments
575
votes
Stop these "5 person premade" missions until you have a proper replacement for Public chat rooms!
Gameplay27 comments
268
votes
I got hit by a Morgana snare at lvl 1
Memes & Games42 comments
233
votes
Ekko Q is visually misleading
Gameplay70 comments
176
votes
Duskblade of Draktharr's slow should be for melee only
Gameplay155 comments
171
votes
Another hidden power of Yasuo that I think is excessive
Gameplay113 comments
147
votes