That’s the same as Custom MMR on hotslogs. That’s for fun only.
Or when brawls existed, brawl specific rankings.
“There’s no need to worry about what rank you are” means it doesn’t work like SL with 2 leagues disparity, but it doesn’t mean that it has no MMR accounted.
We can have a master player with a bronze player, but so does QM and UR.
In my last 18 games, there was an difference of 115 MMR between both team and I’m not even accounting the MMR penalty when stacking, which could be close to the initial ±100 from my OP.
Your last 20 games sum it to 143 MMR difference. The higher number could be explained by 4-5 stacking more often.
And it doesn’t mean there is an MMR. The MMR system in Heroes Profile is decidedly flawed, and the least reliable statistic on the site. For example, let’s look at the values of the most recent ARAM I have on that site (I don’t upload regularly, so that is definitely not the more recent ARAM I have played).
I was in a 3 stack with 2 randoms, the enemy team was all random. One of the people in my group is a current forum poster, and I believe are currently Diamond in SL. The other is a former poster, and is Master in SL. However, it is listing me as being at roughly 2600 MMR with both of them 4-500 points lower. The irony here is they are both better players than I am, I just have more matches uploaded.
The enemy team has an even greater spread with the highest MMR at 2800 and the lowest a thousand rank points below them.
What makes me the most suspicious about “the teams are close in average MMR” is the fact that the average MMR for each team is also the average MMR for HP’s rank assignments in general. I would argue that every single ARAM match will have a average team MMR close to 2500 because that is the overall average for the entire playerbase, so it is simply a relic of how MMR is calculated, nothing indicating an actual in-game algorithm.
2 Likes
OK, so out of a sense of morbid curiosity, I went and looked at their profiles in game, and yes, they are Diamond/Master in SL. So, how about account levels and ARAMs?
My account level is 1961, I have played 586 ARAM matches (note, ARAM match statistics were reset when they made the change from Brawls to ARAM a couple of years ago), with a 51.7% win rate. This suggests that my MMR should be slightly above average, so high Gold at best.
Waterlong is account level 1607, with 181 ARAM matches played, at a 59.1% win rate. This suggests that they should be ranked much higher than I, yet Heroes Profile shows a much lower MMR. Catherina is account level 532, with 101 ARAM matches played, and a 62.4% win rate, and yet her Heroes Profile ARAM MMR is well below average.
I would definitely not use that site as an argument that there is an in-game MMR, just based on these few examples.
1 Like
waterlong only has 21 games uploaded while you have 152 and Catherina has 7 games. The only reliable MMR is yours.
p-value looks scuffed when there aren’t many uploaded games. We can only take into account your MMR. I’d say to track players that have at least 100 games uploaded to make a better assumption.
Could you tell your experience on EU ARAM? Solo queue preferably.
Let me log in and check, but I think only about half of my EU ARAMs were solo queue. Also note that since replays are not saved for Industrial District, Heroes Profile will always be off as in-game stats take those into account, where HP cannot.
Edit: OK, checked EU. For 2020 Season 4, when I first started playing my EU account regularly, I have 26 ARAMs with a whopping, brace yourself, 38.5% win rate. Most of those were solo queue, as I wanted to level heroes I didn’t own for gold. Total overall, I have 70 games at a 48.6% win rate, as I mostly just play ARAMs with friends these days. So in other words, my experience is pretty much the opposite of yours. In some ways this makes sense, as I am an average player in general, especially on heroes I don’t play much other than in ARAM, and on EU I am contending with higher than normal ping.
Nope no mmr in aram. But why should it have ? its just a clown mode to have fun in.
Wrong, every matchmaking process in any game whatsoever has an MMR system. Its just how it works in gaming. (MMR is a concept taken from Chess)
For games like League of Legends you can check : whatismymmr(dot)com
It will tell you for all game modes. Even Aram.
This isn’t the LOL forum, Bubbelz…
3 Likes
There is a section of the community who refuses to believe MMR isn’t used for matchmaking in ARAM games. Even though people can be matched with AI there, they don’t believe it. Even though the Devs stated when it was introduced that there was no MMR in ARAM, they don’t believe it. Even when Hoku links the relevant link to prove their is no MMR in ARAM they don’t believe it.
I don’t believe it. 
8 Likes
The Tooth Fairy was my parents all along.
3 Likes
I am not taking any position about the existence of an MMR system for ARAM. As usual, I will just nit pick, here about « equal footing » statement of the video Hoku linked. The sentence is placed after another one talking about « having heroes unlocking » (I don’t remember the exact sentence). This equal footing could as well be referring to this only and not « equal footing » as a whole (no MMR).
Édit : ok, just rewatched the video and it actually also talks about « not worrying about rank ». This would indeed mean no MMR.
theres no “position” to take over it any more than over the flatness of earth
people who keep insisting there is mmr in a mode that puts literal ai’s on teams are like flat earthers, clueless
1 Like
Well, just because there are AI is not enough. A 1vs1 even with 4 AI teammates could be considered as « fair » for an ELO system. I don’t play enough ARAM and not at the « right » to know about it, but if the system doesn’t put the same number of humans in each side sometimes, it certainly can be considered « unfair » at this point.
third party parsers posting mmr does not mean a system uses mmr. However, since people tend to look for simple answers to things they already want to think, people have a harder time telling the difference between ‘proof’ and confirmation bias.
Lots of games don’t have mmr, so an assertion of a grey-named account that doesn’t know that is doing a better job of indicating what they don’t know, rather than what they do.
Heroesprofile can post ‘mmr’ but since that doesn’t actually do the matching, it doesn’t have any bearing on the matches actually being made. People can script a k-value, rate of change, and run algorithms to post numbers to a name – cuz math isn’t hard, esp when the process can be automated – but that doesn’t mean the host game itself actually uses those. Since parsers see matches formed, and they have to assume the matches are made to be ‘legit’ to a matcher, then the numbers will be evaluated to act as if there is ‘matching’. That is skewed data to confirm an assumption, and not ‘proof’ that it actually exists. A key part of the indication of that skew can be seen in data bias trends esp from ‘season resets’ that don’t have a bearing on particular modes. However, it’s not like people that assert such things actually put in the effort to analyze those quirks, otherwise they would notice the issues, and realize the problem of their assertion.
And that doesn’t happen largely because they’re already convinced ‘they know’ that they don’t bother to look.
However, while there may not be active match-making ratings, that isn’t the same thing as not having any matching priorities. Usually a system will still have priorities to match, but it make not have a ranking system evaluating a value that’s based on the average winrate performance of any given player. There’s probably some ping/region evaluations on the matching, but that is very much not the same thing and I’d like to hope nobody is mistaking ping parameters with with mmr.
But of course that isn’t the case.
No, ARAM/Brawl does not, nor has ever used MMR. It just takes whoever is in the queue and puts them into a match together.
It has always worked like this.
1 Like
On my smurf, EU region, lvl 1.
4 - 0 without a sweat, it’s not even close. I can go on like I did on my main, but I’m bored of stomping bronze mmr on both teams.
Name call me flat earthers, biased or whatever it makes you feel good about your own argument, but I’m yet to be proven otherwise about having a mmr.
Devs said in the past that 4-5 stacks don’t have an advantage, but it’s been proven wrong. 53.6% for 4 stacks, 49% for solo.
A video from 2016 saying that in ARAM, there’s no need to worry about what rank you are isn’t telling that there’s no matchmaking.
AIs filling players in a dead region isn’t telling about the MM. Like this one, it doesn’t count. They’ve even put it equally.
Is there any master players and non OTP that can go cross region on a lvl 1 account and willing to test out the difficulty in ARAM?
I can only bring a short amount of practical info by myself.
A single data point with a very small sample does not constitute evidence.
1 Like
28 - 6 on my smurf in NA.
And yet, my experience was the opposite. So now that we have conflicting anecdotes, what can we conclude?
A Diamond player will always have a much higher win rate than a Gold player if there is no MMR, so these examples are actually showing the opposite of what you are claiming.
4 Likes