For example, if you want to calculate the probability of a streak, you would use a Binomial Distribution to calculate it.
( n choose x ) ( P ) ^ x ( 1 - P ) ^ ( n - x )
( n choose x ) would evaluate to 1 when n = x, and since P and 1-P are exactly the same since it’s a 50% chance it just boils down to the following formula to calculate a streak.
So really, you have like a 1/1000 chance to get a streak of 10, yet, streaks of 10 occur frequently.
The whole “streaks happen” argument without the context of the probability of that streak is fallacious. They do happen, and quite frequently in small amounts, but a streak above 4 is highly unlikely, and a streak of 10 really shouldn’t happen.
Unless of course the player is trolling which will lead to a losestreak no matter what about math or if the player is severely below their rating or learning very quickly which would result in a winstreak.
Mentality plays a big role too, tilted people make more mistakes etc.
Might also be that a loss streak of 5 for example after being tilted could be in a way redeemed as your MMR would tank faster than your rank and then you would face easier opponents just because you had a bad day and a lucky winstreak countered by your MMR rising faster and then facing better enemies because its bloated without an actual rise in skill.
Not forced as much as just bad MM or too rash movements in MMR.
Sure, link your hotslogs/upload the replay of the game in question.
Individual games are kinda meh to evaluate in the spirit of discussion as we are talking about an concept that rules after mass amount of games and such.
I haven’t use Hotslogs in years but it’s fair to say that we did not perform well with these hard-counters in the matches. These stats don’t mean anything in correlation to an uneven team-distribution but this should give you a sense what we have been up to.
Games in the right look pretty rough, like hammer vs ming, Nazeebo and chromie.
Games on the left could go any way really, depending on the map and such as well.
Personally i stopped playing QM unless i really want to try a new hero in PVP before going ranked with them but otherwise i cant really enjoy chaotic comps with no healer or a tank or either.
Then you will likely understand why people say that the matchmaker is bad, don’t you?
Everyone of these games were bad. In one we had a Account LV50 player who was genuine new and had no idea what he did. Another one was simply decided by matchup, in another we could have won but we were always outnumbered.
And of course the game thinks then you belong down there. If I don’t go in and try at least as a support, then they would have died sooner. And on top of that, we had in three of the four matches an Abathur who couldn’t perform as well. Most of these people don’t even collect globes next to them. This is the kind of quality you’re facing there as a high level player.
I’m pretty sure we got mmr-averaged out so the games could be found within 5 secs of QM-search.
We have no other choice here in our region. UR takes around 5-10 mins before we find a game, QM is the only mode which will likely find games <2mins.
Well you should know that i think that as i have said it in this thread too multiple times.
UD, QM and TL use MMR averaging and its at its worst with parties being involved.
I got some friends i have a 90% wr with because their MMR in QM is so low that i face people who are in Silver in HL and such and as i mainly play assassins its a walk in the park to win these games.
Im in Diamond atm and climbed there from Silver in HL.
Diamonds/ low Masters are what i usually play with and against in the other modes too.
heh match making for HOTS has been the biggest trash since it launched hell tbh blizzard has always been absolute trash when it comes to match making and ranking systems.
They .must work.fast to.compelete cross.server.EU, US program.
For 51617172 times, players cant play in every game with his best way!.
And maybe is possible to match with different league if is not exist online players with ne same devisione or league.