Experiment Proving Getting Reported Places you on Forced Loss Streaks

The game is dead people are leaving

There might be some effect going on that will alter win rates that we don’t fully understand, but I just can’t believe it’s something that’s so direct and intentional as punishment for being reported.

Really? This is news to me. How do you know this?

Forced loss streaks are no doubt real, but I believe them to be an unintentional side effect of the terrible match maker.

The match maker tries to create even matches where both teams have 50% probability of winning.

Where do you think it gets this data from?
Winrate (if you don’t believe me, then explain from where else).

So someone on a massive win streak or massively positive win rate, gets put with donkeys to even things out.
This is a fact and I see it all the time.

The better you are, the more donkeys you get.

I suspect this is what happened to you, ranking up a new account, winning the vast majority, the MM gives you donkeys for winning so much.
If it didn’t then the team would be massively imbalanced and have dramatically greater than 50% probability to win.

It’s a bad matchmaker, how Bliz can’t are this from the thousands of posts made in the last 3 years baffles me.

4 Likes

They’ve expressly stated winrate is not used for matchmaking purposes:

At no point in MMR calculations do we look at your win/loss ratio and win/loss ratio is never used to determine who to match you with or against.

If you want a good thread that has a ton of citations and sources to posts detailing the information we’ve been handed on the matchmaker, check out Kaawumba’s thread (Even if you don’t agree with his conclusions, it’s a good source due to the citations):

2 Likes

I’m curious, were you paying attention to toxicity levels? I feel like it’s possible there’s a toxicity que. If your whole team is toxic, you are more likely to get rolled, agreed?

1 Like

Just because you have in the past doesn’t mean you will always be able to in the same amount of time in the future

Yea… even if this is true it doesn’t mean anything honestly. You could have been up against Smurfs that were better than you for instance.

That is wholly insufficient for data. No video? No screenshot?
Easily disregarded

Doesn’t mean you will in the time that you are aiming for. Just because that tier is easy for you doesn’t mean you won’t get derankers on the enemy team that are better than you. There are factors you are not considering that are important

Again, easily disregarded as we have no way to actually see that this is the case. You could, easily mind you, just make up data.

I don’t believe the matches are rigged and I will not do this, sorry.

I could get banned, muted, or otherwise have action taken on my account.

How about you do it 30 more times and actually show us the video and pictures.

Then you may have something of a case.

Until then, this is just hearsay and it is disregarded easily

1 Like

One data point means nothing. Just because you played a bunch of games doesn’t make you more than one person.

I would like to point out, this is not a scientifically conclusive experiment. There are many, many factors that could cause this. Along with the fact that it is only a single test. For an experiment it has to repeatedly be conclusive and for others to be able to recreate it. A single test result is not conclusive

2 Likes

Wow, thank you, you must be very knowledgeable in this Overwatch game.

Sorry, I couldn’t help myself, that was English humour (commonly known as sarcasm).

I’m 34, a senior solution architect (software consultant), working for one of the largest consultancies in the world, mostly focusing on financial and retail, I’ve a master degree.

All things told, I’m no dummy.

Explain to me in simple English, like a child would understand, how you would calculate win probability?

If indeed you chose to NOT look at winrate, how would you calculate it?

It certain doesn’t look at the team role distribution to see if it’s a viable team, otherwise 4 mercy main teams would never be a thing.

I can assure you, there are no other measures that they could possibly use.

Don’t believe a word they write, as half the time the person writing it actually has no clue, the other half they’re intentionally giving misinformation, as otherwise they’d be rumbled.

Think about it before you come back with more quotes, ask yourself how else could it be calculated to 50% win probability.

3 Likes

Might want to tone down that superiority complex Mr. Smart Guy. Hubris has been the undoing of many intelligent people.

2 words - statistical performances.

If we have a team-based fruit gathering contest where Player A averages 3 fruits with a 70% win rate, B averages 7 fruits, with a 50% win rate, C and D both average 5 fruits with a 30% and 50% win rate, you create the match based upon the expected gathering of fruit (where each side is expected to produce 10) instead of balancing win rates (which produces a team expected to produce 12 going against a team expected to produce 8).

It’s a gross oversimplification, but that’s what you asked for.

If you want to go the full-blown conspiracy route, more power to you. We can’t have a reasonable discussion, though. The moment you reject everything that’s in Blue text is the moment anything that remotely resembles a reasonable premise has to be considered an actual possibility. It is to say “We know nothing, and therefore, it could be virtually anything”.

4 Likes

“Statistical performance” is your response, and within that response you refer to winrate constantly.

Funny, aren’t you the very person which said winrate had nothing to do with it.

Why would it look at pure stats, with no context?

Low deaths could just mean you’re doing nothing, high elims from stat padding, high damage from playing junkrat etc:

Stats alone prove nothing, certainly do not prove that you are able to win games.

Do you know what does prove you are able to win games though? Actually winning games and this is reflected on winrate.

Sure they may look at other stats, I’ve no doubt, but to not be looking at winrate they would have no ability to even calculate a crude win probability.

I win most of my games from shot calling, does that mean my stats are insane? No. But am I winning games? Yes. Therefore the probability that I’ll win another game is high.

You talk the talk, but unfortunately sir, you cannot walk the walk and clearly you have very limited experience in data analysis and statistics.

3 Likes

You requested a 1st grade answer, and then complain the response is too simple in nature and create a strawman. It’s almost cute.

I never claimed winrate doesn’t factor into MMR. It’s self-evident that it does (as it climbs when you win and drops when you lose). My claim is that winrate isn’t used when creating matches. This prevents things from happening where a 3200 player on a loss streak isn’t valued significantly lower than a 2600 player on a win streak when they both end up at 2900 MMR and placed in the same match.

Unfortunately for you, even if I were to concede you were more well-versed with statistical analytics, it wouldn’t make a difference when discussing the matchmaker. Neither of us have access to the data or information that would be required to form objective opinions.

I provided a source from someone that does have access to that information that plainly stated they don’t use winrates to form matches. That’s significantly firmer ground to stand on than someone that comes in and says “They’re all lying. I’m super smart. So, I can totally tell.”

Edit: Additionally, I’m cool with agreeing to disagree. Our side conversation on winrates in the matchmaking algorithm are a bit off-topic for the thread’s subject.

Apologies, S23. I get distracted, easily.

3 Likes

In addition he knows what is being tested so he will unconsciously affect the results of the experiment. Double-blind experiments are a necessity

3 Likes

See thread.

Agreed it is insufficient. Why not join the test? You can do this too!

Oh, you won’t. See:

And then the same excuses of “I might get banned” that I stomped on earlier. Read the thread people, before making claims that are already debunked.

Join in?

Join in?

Anyone else noticing the pattern that I claimed would happen right at the start of the thread?

Why do people misunderstand and think the thread is about a punishment queue to force losses? It is about some “low priority queue” style structure where you get put with other reported people, until you’ve played enough games without being reported. Super common in games, and could have been introduced to this one in the last year or so. The idea is to force trolls to play with trolls. But they will still play against normal players, hence why it is likely you will get stomped.

I actually am not convinced this happens, but I’ll try it out rather than attack the idea.

1 Like

Says “see thread” when there is no evidence in the thread.

Lovely

I went through the thread and literally nothing you said offers any proof that your claim is true. So “stomped on” is entirely inaccurate.

Just goes to show that if you have to exaggerate how well you defeated an argument, then you probably didn’t defeat it very well or at all. Much like with you :smiley:

Because it wouldn’t be blind and, as a result, my bias would unconsciously affect the result of the test. I will join the test when it actually is scientific, not just an accumulation of unsubstantiated anecdotes.

You’re going to get nowhere insulting me. Firstly, I’m not “in denial”. You have simply not shown the truth of your claims. I also do not want to get reported over and over again. I also won’t try because of the reasons I mentioned above (not being blind, not being scientific)

Please show me where you “stomped on” the concern of getting banned, because I don’t see it.


I looked and you simply don’t “stomp on” anything actually. Please don’t claim you have “debunked” anything when you have, literally and objectively, done nothing of the sort. It only makes you look worse than you already do. Thanks


I suggest you learn the absolute basics of how to perform an experiment and what actually constitutes as proof or evidence. It will help you in the long run.

Even if you “try it out” and your “try” shows that it is the case it would mean nothing. You are not performing this experiment blind and, as a necessary consequence, your bias will absolutely change the result of the experiment and/or how you interpret the result.

Also, by expressing doubt and pointing out your flaws in logic I am not attacking the idea.

Please study what science actually is.

1 Like

If you read my initial comment, I stated that I believe it may be the case that there are low priority ques. I just intended to point out the major flaw in this study and how it isn’t enough evidence to be accurate. It’s not my project, I don’t have to create evidence for them.

3 Likes

Been there done that. Also studied logic. You might want to try that since obviously you haven’t. You couldn’t even work out which part I said stomped about and misquoted it heavily.

Cool, then don’t claim the idea is bad because of lack of data?

I never said it was a bad idea, just that there is not enough evidence to be conclusive.

2 Likes