Ive made my claim. You seem to have overlooked it in favor of arguing another point.
Its not fair that people playing ranked get paired against higher ranked players who have lower unranked mmr. Especially since they are not risking mmr.
and as I’ve already said, the way the MMR system works, it doesn’t matter in the long run because you will not lose many points to someone who is much higher ranked than you.
“It doesnt really matter because its only unfair a little bit”
Variation is to be expected no matter what, its how the system works and you will always be losing/gaining points despite what you consider “fair”.
That’s more or less what I told him ; if people do play sufficiently in both modes . Aside of matchmaking logics, the other point in favor of that is that non freelosing smurf accounts in different zones tend to get quite close in MMR (for example we’ve got a M1 in our community who is around 5.1-5.2K MMR in EU… and about the same in NA).
I merely report the ratio between the different types of smurfs confirmed at my level, and regular players. Those are small scale, unpretentious stats (a mere 300 players sample at the moment), but they are also the only ones on the subject yet. The thread aims at providing advices to regular players about how to deal with/spot smurfs ; and at monitoring the extent of the phenomenon. I also intend to communicate those to the devs for the next community feedbacks.
There have been several occurrences suggesting that the devs do read Community feedbacks, and some General Discussion threads on the US forums. I myself have been listened to a couple of times (regarding the Reaper nerf, the Marauder de-nerf, and the Thor’s AA splash nerf → single AA buff). So, even if it’s buried under torrents of useless balance threads whines, the probability of the devs retaining some bits of the US forums isn’t zero.
Yet, you’re right, TeamLiquid forums have also been source of useful data for the community, and some of their bugs reports have led to balance adjustment. I may post there as well in the future. Yet, I also believe that having some constructive users on Battlenet forums isn’t a bad thing for the community. 
Theres actually quite a lot of match data available through websites like for example aligulac or accessing the sc2 API (not 100% sure about this one).
It would actually be pretty easy to detect smurfs by examining players with huge swings in MMR that tend to go on win/lose streaks.
1 Like
As far as I know Aligulac only applies to highlevel players, whereas freelosing smurfing seems on the contrary to be more of an intermediate to low leagues issue.
I also have some doubts on the devs allowing me to stress the API system by trying to detect MMR swings as that would require regular full database scans. And in fact, the smurfs who do all their freeloses in one go are quite minority (around 1.5%) compared to those who freelose regularly (around 20%) .
Yet, even if they’re a minority, data applied to the whole database would be precious. I may contact RankedFTW to see if detecting those would be doable. Thanks for the idea, Cheezecake. 
No it wouldn’t, you could just get a large random sample of players and then examine their match history and use this proportion to come up with a confidence interval.
But first you have to get rid of your god-awful way of attempting to classify smurfs. If you can get access to players MMR from game to game then it would be really easy to detect them from how i explained above.
That’s not what you suggested earlier. It’s closer to what I’m doing now, but in an automated way. That would be great though, but I’m quite unsure players’ history is currently accessible from webAPI (its not from RFTW or battlenet profiles), so I’m not sure of the feasibility of the thing.
If you have a directly applicable way of doing so I’m all ears.
As for classification, if we get past the title of my thread, it’s just simply necessary, as otherwise people would confuse the original purpose of highlevel smurfing, with the ones plaguing intermediate to low levels.
The other benefit of classifying them is that it may provide insight on the reasons behind this behavior, which might be interesting to deal with the issue, including from a game designer’s point of view. The motivation is not the same when a player decides to freelose only one match-up, or just freelose all to get to a certain league.
Well…I only play twtenty five ranked matches or less every season. I just want to illume my master emblem or grandmaster emblem. Emblem is beautiful. Sometimes screen shot my emblem to post in my discussion group. And enjoy it ,and over. If I want to challenge myself such as improve my multiline-handle, I will chose to find an oppoent in unranked match. If my oppoent is a pro player who has 6000 or higher MMR and I get defeated, I would consider the reason and think it’s course. If I won the match, I would think he lost to me on purpose unless I won him over three times. If I won him over three times, I will check all of the replays, consider his faults and my faults. Of course, when I lose, I check the replays too becuase I want to find my faults. Becuase I want to draw a perfect circle and perfect squre at the same time by my both hands.
And I didn’t play Starcraft 2 rencently time becuase it’s hard for me to win in PVT late-games. It’s normally in Starcraft 2 so I try other Blizzard games: come back to Azeroth and play as Ana in hero of storm.
By the way, play as Ana is really fun, save my teammates give me more achievements than beat my oppoent.
That’s an episode:
I match with my yesterday teammates again.
He ask me:
How did you do that? I watch last night replay, your screen is Shaking! Seems like you focus on three location at the same time, you were considering what happen in three locations?
I answer: Yes. I have played Starcraft 2 and got grandmaster.
When I was young as you, it’s just normal----Ana quote.
(Oh, dears, it’s cool.)
what your doing now isn’t at all statistically sound, its incredibly biased and as I told you before, you cannot come to any conclusions with your data.
It would be biased if I generalized my observation to the whole database. If I specify the limited range of it, then it just becomes a limited range report. Which AFAIK, is the only work done on the subject yet. And you know the saying : in the land of the blind, one-eyed are kings.
Now, I’m not sure players’ history is accessible from the API, nor if it could be done to retrieve its data automatically. But there might be people able to help me out on this, and even if the access was manual I’d be willing to invest some time into it. So I’m going to consider that possibility, thanks for suggesting.
Oh, dear, he is right. In other industries such as banking industry, we tried to establish quantitative models for all customers such as credit risk+ model ( The quantitative models for helping us to consider whether customers have enough money to repay their debt on repayment date. But we still have many non-performing loans.) That’s just a reference, we still try to directly ask our customers what they need, and think about what they really need.
Another example: High frequency automatic trading, according to the history data. It actually bankrupted a lot of people. They trust they can quantifying the entire market, is it a wise choice?
1 Like
« The more you know ». When there is much at stake, knowing more about the whereabouts can only make your decision better. The thing is that by that, you should also know what you DON’T know, otherwise the decisions coming from it may indeed not be wise. That’s the parallel I would make with trading.
In my case however, the stakes are very limited. So it’s reasonable if the data gathered is of limited range/quality as well. 
No. I just want to told you. Don’t try to quantify the entire market. A lot of idiots have tried. You are genius, not idiot.
1 Like
I’m certainly no genius, but there’s truth into your words as smurfing may be more frequent at certain MMR ranges than in the whole thing. I’ll take in consideration both point of view though. Thanks for the input (often rich in crispy anecdotes btw 
), Eternity.
This has been the most proper conversation ive seen in a while on these forums.
1 Like
** Black swan event** check it. You can understand. When it happned, we find the best method to solve it as soon as possbile. Try to fix the worst and then you will gain additional market share. Don’t try to quantify the entire market and control it. If the smurfing of certain MMR ranges lead to the whole thing goes worse, fix it as soon as possbile. if not, you needn’t try to fix it, you just waste your time because you can’t control the entire market. When you try to control the entire market, you are creating the new problems to ask for troubles.
Well…I just feel boring, hope can help you.
You are biased because you are only taking your own matches, and not a randomized sample from whatever range you are talking about.
This wasn’t an issue of trying to quantify the entire market, this is an issue of trying to use a model out of data, in other words if you have a model then you only assume it works within the range of whatever data you have, you can’t extrapolate.
Within my MMR range, are my opponents selected in a predefined, unchanging, unvarying manner ?