How many win's in QM before it forces a loss?

General Discussion
41

Long story short, the intent of said removal was to remove the self fulfilling prophecy bias.
Stacks are silly because some are random friends, exactly as coordinated as random teams, while others use voice chat, now each other’s style. The MM cannot know that, so it plasters an MMR adjustment on stacks.

It’s still easy to see that you’re against a party (due to rainbow ranks) but you don’t exactly know, so you only have to worry as much as you would in a quick match (knowing it’s rainbow as well): be mindful of the high and exploit the low. I guess there is a point where you go to level 2 and notice the team dynamics.

Anyway, without a bias, the devs can shoot for appropriate adjustments.

I’m regularly playing against stacks and I’m regularly winning those. Arguably my rank is about the middle so my enemies are unlikely to be coordinated. Maybe that adjustment should be rank specific. Maybe it is.

42

If 50/50 is real why am I 66% win rate in SL?

If 50/50 is real why I’m getting 60% win rate QM players.

:joy::ok_hand:

1 Like
43

You’re right, it doesn’t force wins or losses. I’ve seen people with lifetime win rates of 70% and others with 30% in QM. There are too troll players I’ve seen who try to lose every game by sabotage. They may deliberately walk into to enemy towers and die as often as possible, but even they can get a win despite their efforts to lose. One of those trolls I saw had a 12% win rate over a couple of hundred games.

I have occasionally played too many games in a row and I think this is in part a perception issue. You may win more games while your mind is fresh, so after six games you might feel focused, but your brain is probably wandering to other matters. This is where you might start to lose and you will have a streak of them if you don’t quit for the day.

The number depends on the individual, but a forced win rate isn’t a thing.

4 Likes
44

nothing as sinister as this, it’s just your MMR increasing til you are playing against players far better then you.

3 Likes
45

The average player is average at the game.

The matching system tries to average mmr to create an average probability of ‘fair’ play between the teams. So the average average is averaged.

Player populace (mmr distribution) falls into certain peaks were some values are more common than others, which influences where an ‘average’ of the mmr is going to fall. If a team has an average of 2000 mmr, than that might be a composition of: 1800,1900,2000,2100,2200 rather than 1980,1990,2000,2010,2020.

Before players ramp up their ‘win streak’ they may be in the same averaged block, but they’re matched at the lower end of the mmr (1800 in this case) so there are ‘better’ allies in their group that can be more cooperative to play alongside. Player wins some games, mmr increases, but they’re still in the same relative average mmr block (2000 in this case)

After wins pile up, their personal deviance from the average has increased, but the matching average is still relatively the same. Their game is filled with “worse allies” relative to the previous few games they were playing.

Now they’re the 2200 in the spread of 1800,1900,2000,2100, but they are the one above the others in that match For the player experience this can feel like the “forced loss” scenario while just being the 'working as intended" model for the mmr averages.

Part of why some suggest taking a break, or queuing as a different hero is to then shake up the systems expected average it has for matching. If you’ve already played several game, the matching has an ‘average’ it’s looking to make, so its not uncommon for players to end up in the same queue as players they may have already played with/against because they’ll all still in the same relative average mmr block. So taking a break, or queueing as heroes that adjust the mmr averaging then shift the block a player is matched against.

So… basic math is pretty much why the system can seem ‘forced/broken’ and “working as intended” at the same time. However, instead of pointing out how averages can work (and realizing the extent of player populace in actual ‘skill’) people usually draw up boogie man stories and have to profess the universe/blizzard/activision/whatever is personally out to get them. It’s much easier to blame something as some anthropomorphic evil impossibility (that would lead to better matching if it were true) than to do some arithmetic.

Cuz afterall, once school is out, math is the greatest evil of them all!

edit note: I am referring to mean, median, and mode in this post. “Average” doesn’t just refer to one thing.

5 Likes
46

Idk ask the community. I don’t care to post my opinions on rather its real or not just point out the communities blind beliefs.

2 Likes
47
  1. Humans are not performing on a constant level.
  2. The MMR is not a static number, everyone falls and climbs all the time, after every single match.
  3. QM, SL and UD has their own MMR, so one’s performance in one mode means nothing to an other.
  4. ARAM and vsAI has no MMR.
  5. Teams form an average MMR and that’s what matched.
  6. Smurfs exist.
  7. The average player is not good at the game.
  8. You cannot statistically win for the end of time.
  9. If you play on your skill lvl (you’re average on your MMR) you probably win and lose the same amount since you can’t push the game in either direction since you’re not better to carry, nor worse to drag down.

Therefore, there’s no need for a System that forces losses, nor could one be made.
Which seems to be the case, since there are multiple ppl with high positive winrates (to not use the low winrate accounts cuz “losing is easy”).
After all, how can ppl climb if there’s a force that should stop them?
And if the skilled players can overcome a hypothetical “force” that want them to lose, why ppl think it exists in the first place?

5 Likes
48

there is no matchmaking in AI

this isnt the flat earth society, opinions dont have the same value as facts, people keep repeating “forced 50%” but it isnt real

49

You should win most games. As I said, Hero composition and map are also possible factors. I’ve gotten Abathur, Nova, Valeera, Samuro and me as Zeratul before, and though I have won such scenarios, sometimes is just impossible because the other team has a proper composition.
But, since there’s no way to prove hero and map selection aren’t completely random, my point stand still that 50/50 is not a thing in AI. You could farm beginner for ever, for example. Which is what many do to get lots of XP.

Ohhohohohoh, I have suffered this many times, most often than not when I go with some hero that lacks self-sustain (hello Maiev). But for AIs it doesn’t seem to make that much of a difference? It depends on the healer, but healers such as Li Li hold on too much on their heals, and in 5v5 AI don’t usually have a great impact so they can still fight somewhat evenly.

This is all, of course, just my observation.

50

Hasn’t the 50% forced just become the new “Glass half empty” way of thinking?

2 Likes
51

Only when you want to deflect from the issue. The stigma gloves have been off a long time ago.