What's the logic behind the LP system?
So same division. You lose 5 games in a row. If you win 5 games in a row you have 16+17+19+19+20 (91)
So same division. You lose 1 game, you win one game. Take the same number of games (10), overall same 5 game win, 5 game loss; but now you get 19+19+21+19+20 (98) .
That's 7 more LP. But extend this to higher losing streaks and not followed by win steaks .... Voila. You get huge differences. Like now, I'm after a losing streak and get 16-17 / win if I win 2 lose 1.
This right here is the proof the system is skewed to throw you down unless you are very good in a division to get constant win streak (basically a smurf)
A good ELO system has the same points over the N number of games no matter the way the win/loss are mapped
It amazes me how people are content with this. Who the hell was the "brain" behind this system?
Clarifications
I know the MMR is hidden. But that's not the point. The values are for the same sample size: 10 games with 5 wins and 5 losses. The difference is win distribution. In ANY decent MMR system, WIN DISTRIBUTION DOES NOT MATTER**! . Because right now, a losing streak is more punishing than a winning streak. What does this mean? Since for any non-smurf player losing streaks are something unavoidable, means the climb is skewed backwards, to properly climb you either need big win streaks or over 52% percent win rate on a huge number of games, or you can't climb.
What does this mean? tl;dr current system means you either need constant big win streaks or a huge number of games to climb. Basically first part is you are a smurf, second part is for normal players. If you are not a smurf, good luck proper climbing in less than 500 games.