Quote:

**TheBlackJeff**:

Unless the conditions I already stated have happened (leaver/afk or at least one player who has a significantly higher or lower elo/mmr compared to their true skill) then the "higher chance" you are speaking about is not statistically relevant.

You seem to have trouble grasping this. The matchmaker does not determine win or loss except in the above scenarios. (and even then it only skews the odds) It does not randomly spit out teams that are 1200s players vs 1800s players.

The most significant factor in an individual's win/loss ratio is their own play relative to their peers. If they are indeed more skilled and play better than others in their current elo range, they will cause their team to be the "good team" more often than not, and rise in elo/mmr.

This is not random. If you really do belong in a significantly higher elo, then playing with 4 members of this current *lower* elo on your team and 5 on the opposite team is not going to yield teams that are "*bad enough for him to be unable to carry, enough times, that it prevents him from going over a 50% win rate, halting his advancement to his true elo.*" Arguing that is simply nonsensical.

for most people, you're correct.

but for a very small percentage, you are incorrect.

In a game with millions, a very small percentage is thousands.

It is very POSSIBLE that someone can in fact "

*bad enough for him to be unable to carry, enough times, that it prevents him from going over a 50% win rate, halting his advancement to his true elo.*
Probable? Not so much. But even the slightest probability in a sample size of millions is a SIGNIFICANTLY large enough number.

ALSO. Match making doesn't match teams up against each other with perfectly equal ELO's. Meaning:

Team A could have a combined ELO of 6123

Team B could have a combined ELO of 5901

Match making doesn't wait until it's perfect. It puts people together that are reasonably close as best it can, as fast it can. It's not PERFECT.

And also, that ELO for team B's members could repeatedly be higher than their actual skill, increasing the POSSIBILITY that elo hell exists.

Improbable, yes.

Impossible, NO NO NO NO NO.

in a sample size of MILLIONS, even a .01% chance equals THOUSANDS.