3k MMR? It’s not low. I mean, queues are getting decently long and I start seeing the same people in each game. Not that I think it’s so hard to get to, but the pool of players really starts to thin out. When you start inserting players in that are substantially below that, it just makes those matches that much less fun. It’s really easy to see who understands what’s going on and show doesn’t.
Yea, agree. There is something working in the background it seems like. As soon as I start losing, I get put in games that are absolute losses from the start. If it didn’t keep happening, i’d dismiss it, but it’s really consistent.
I started on one of my shams early and hit 2.7 mmr very quickly. I don’t know if it was week two or three, but regardless, MMR was getting high pretty early in the season.
My mage was ranked 50ish on the leader board and I was around middle of the pack on the first page for games played. The played games vs CR is all over the place.
Yes, the inflation started right around the time most of the top of the ladder stopped queueing and swapped to alts on Monday night of week 1 (I know because I was there; the 2.2kish MMR queues went from 3-5 minute pop to not popping for hours because there were no longer enough players left at the rating to fill a lobby; I didn’t get another lobby on this character until some of those alts got to around 2k)
This hunter was in the top 3 for surv for the first 2 weeks of the season before I had 3-4 back to back disconnects and tanked, and both are currently 24/25 on the survival leaderboard. This character’s ghost on Drustvar from before I server transferred it is still in second place for games played as surv at 729 (real total is now 928).
Played games vs CR is absolutely not “all over the place.” Not a single person on the leaderboard with a winrate over 70% has more than 250 games won. Nobody with more than 500 wins has a winrate above 57%. That’s indicative of how many games that Blitz requires to knock down the nails sticking up. That rogue with an 80% winrate (102-25) has a zero percent chance of maintaining that winrate through 500 games without foul play.
The only thing that’s really up to speculation from what I said is what impact duo queue, particularly early on, had in creating these equations. The impact of duo queue and tank lobbies is certainly non-zero, but I have no idea how you would quantify the impact except to remove those features for a season and see what happens (unlikely, ofc)
Just looking at frost mages atm on the first page, you have a mage with 1211 games played with 643 wins vs 101 played with 61. Thats 10x difference. How are you going to argue that isn’t all over the place unless we’re talking about different things. While that’s on the extreme end, plenty of 3x to 5x differences.
I don’t think the duo queue is having any impact as you suggest, though I’d be glad to see it more broken down. I get why they hard capped CR gains and I understand why there is such a massive variance in MMR in each match. In the end, the wider the range the faster we are going to get queues. The main point I was trying to make here is that needs to be tightened up.
They could still make games for people with low CRs and high MMR, but mixing the two together is really taking away from the game, regardless of how we got here. To that, how the system is averaging MMR is putting players on losing streaks that ALSO hurts the players in their group. Low CR players are getting tossed into higher MMR games and they just get stomped. Doesn’t always happen, but it does enough that there are plenty of posts about it.
If those items were addressed, I think blitz would be magnitudes better and maybe even a more serious bracket.
I’m explaining to you that nobody is at their “real” rating unless they’ve played hundreds/thousands of games and flattened their own winrate curve, which is rare.
I’m not defending the current design, I’m explaining why it looks the way it does. The community doesn’t play enough for it to function.
It requires so many games to flatten that curve due to the low point gain cap. The low point gain cap was implemented because of historic wintrading in RBGs. Duo queue is susceptible to the same abuse if points aren’t locked so low, because low and high MMR can queue together.
I hope that better helps you understand what I’m trying to explain here. It’s a complex issue rooted partly in community behavior, kind of like how the “experienced” side of the ladder is more prone to giving up, crashing out, and getting banned for language.
I’m against deflating the bracket too much as you’ll just kill it like with 3s and RSS
I am in favor though of them removing the cap on CR gains after 1600 to make up for said inflation. If someone is 1700 and is winning constantly in 2600-2800 mmr lobbies then let them zoom up to that CR since it’s obvious that’s where they should be playing at.
Little last note here, but the reason why we’re seeing 1600 heals in these lobbies is literally due to either not queuing at prime time which means less heals in the pool.
Solution: Create sub system that when people take a que if someone is way out of their depth / leaves midgame the penalty is nullified, but said person also loses a good chunk of MMR to get them out of said lobbies quickly. Yes people will potentially abuse it, but if they do just DQ them from getting rewards for that season /shrug
1-0 from a newly fresh toon with a 2.5k friend puts you at ~3k atm, which you can very easily do with another high mmr toon that you duo queue with. You can actually leap frog new toons to 2700ish cr with someone who has 2300 mmr.
But most of us would recognize that will never happen for some players. That’s true for 3s, shuffles, etc. What do you consider flat, a perfect 50/50? ‘New’ players are always being brought in and brought up. They feed on the players below them and that carries up the ladder. Instead of reaching 50/50, they simply stop gaining points.
I gotcha, but I still think there are plenty of solutions that could remedy the items i brought up. Be it through strict MMR control, using more data points to decide MMR, etc.
I don’t think it is though. These healers are in the correct MMR range. The priest in almost full greens was 2800. They had won a bunch of games. The problem is you can go from 2100 to 2800 in 2-3 matches.
Also, this is during peak times and it’s not exclusive to healers. I have seen players as low as 1100cr in a 3k match. Something is off there if you ask me.
Would be nice if they just explained the whole damn system, but instead we use the military everything is a secret so people have to figure it out. I get people will game it, but they game it anyway just with less knowledge spread around. Share what you can on it blizzard and keep some a secret sauce.
Right nit feels like it is partially manipulated to keep queue times down even with worse matchmaking.
That goes back to the community issue thing…
There’s two groups of people that will never play enough games to equalize their MMR and CR;
Group 1) they don’t have enough spare time to play that many games, a very understandable situation
Group 2) their ego is fueled by the scoreboard saying their MMR is 3400 and their winrate is 65%, but they subconciously know very well they’re going to eventually hit a lose-streak before making their CR and MMR match, so they only queue up at the crack of dawn when 5-8 of their friends are all in queue.
Group 2 kinda ruins that situation for Group 1, don’t they? Unfortunately, Group 2’s behavior is increasingly common the higher up the ladder you go; the lack of games played shifts from being about free time to invest to trying to match free time up to other players.
A certain warrior streamer with an attitude openly admits to this behavior, since he doesn’t think there’s anything “wrong” with it, as it’s an adaptation of long-standing RBG community behavior with regard to dodging and syncing.
I love Blitz, personally, but some of the flaws it inherited via the community have no real fix (although things like duo queue do seem to make those flaws more pronounced)
The best systems design in gaming will never be a match for the degenerate and exploitative behavior of gamers who view that system as a hindrance to their being viewed as “the best” rather than as an environment for them to compete with one another. Streamers like to run their own little tournaments outside the ladder for the same reason. No pesky randos to ruin their attention-seeking.
I’m not sure why this is an issue relative to the items presented. The majority of players play a very small amount of games and often times end on a negative win/loss ratio. Those losses translate to wins for the next guy and that process continues up the ladder. For every player with a 60/40 win/loss there are 10x that on the opposite end. I don’t think people realize how low the floor actually is. Players don’t need to play until their MMR/CR equalize because there is almost always people feeding the bottom.
There are always fixes. This isn’t a community problem. It’s a problem of how MMR is currently working.
In the issue I’m presenting here, simply tightening up the acceptable range of difference between a player’s MMR and CR would go a long way. To that, they can simply create matches with players who have a high MMR with a low CR and match them with other like players. Queue times might increase during some periods, but most likely the impact would be minimal and the quality of matches would go up substantially.
The way it currently works is itself a reaction to the ways the community is known to have exploited the system in RBGs.
It takes a large amount of games to equalize MMR and CR because CR gains are locked at a substantially lower rate than MMR gains are.
Unlocking the CR gains reintroduces the exploit from RBGs. Locking the MMR gains reduces player interest because a substantial amount of players only play as long as there is a number on the screen that keeps increasing, even if it’s just MMR.
Last time we heard that line, when they tried to hotfix lobbies being full of the same class/spec, queue times increased exponentially for many players.
I don’t think there are plenty of ways to fix this problem. Not every problem has a solution, and neither you nor I have a full picture of the equations currently being used in lobby formation. For example, how do you factor in same class/spec alts into your prospective high MMR/low CR segregated lobbies? Will that become the new form of twinking? It’s already a strategy for minimizing losses, after all.
The only winning move is to play because you enjoy the bracket. It will never be “perfected” any more than 3s has been.
The target audience for Blitz is people who want to play a lot of rated battlegrounds, but were previously unable to get queues due to the stagnancy and gatekeeping of the RBG scene. Largely, that’s the former yolo-bracket players. Expect no more perfection than they did and you’ll have a lot more fun with it.
I disagree, regardless, this isn’t something that can’t be fixed.
There is a fairly large pool of players queueing blitz now. I guarantee you everyone would rather have games that are more balanced then constantly dealing with games of luck.
The MMR system can be tweaked in a 1,000 ways. This isn’t even a complicated issue.
They have patents on the MMR system. I looked at them sometime ago. In the same way they have adjusted how MMR works in each bracket, they can surely adjust it to minimize the outlier matches.
No one is asking for perfect, just something a bit better than what we have now. I don’t think I should be having matches where healers are apologizing because they have no gear and the system placed them in that match.
You didn’t lose that game because the healer had no gear. You lost that game because that healer had no experience, which was amplified by the lack of gear. I can just as easily cite examples of people playing higher MMR/CR games in subpar gear. My alt surpassed my main’s rating in warmode gear initially.
And what will be the knockdown effects of those tweaks?
Your proposed change of segregated CR/MMR lobbies would impact the early morning bracket, which is already a bracket plagued with a lot of shady behavior that is enabled by the very slim amount of players in queue at those hours. Did you consider wintrading at Oceanic slim-queue hours in your proposed change? Because it’s absolutely a factor.
(apologies for the double-post, usually the forum system collapses these into one post together but I seem to have confused it)
3k is low when the highest lobby hit 4k last week. Same gap between 3k and 2k, right?
We’re at the end of the season, you’ll see many of the same people who are also plateauing at that MMR over a couple of games.
Agreed, but CR means nothing. I’d rather have someone at 0 CR who is good than someone who has the CR relative to the MMR theyre in and bad. There’s plenty of completely horrendous players all the way up to 38 MMR.
Ratings & MMR mean nothing in the grand scheme of things.
You’re 28~ in that screenshot, in a 3k MMR lobby - wouldn’t be much of a stretch for someone to complain they’re getting people 200 CR lower than the MMR of the game.
People not so long ago used to not invite if you were that far from the required CR on the group in LFG/Oqueue.
This - but even then, there’s much win/lose streak wave riding which is disingenuous.
I feel like that Group 2 scenario happens less the longer the season goes.
People queue alt DPS specs on classes like Rogue/Mage/Warlock or just make second characters, either different classes or the same spec so they just play both into R1 range.
