No, there is no forced 50/50 – but it might look like that
There is however and performance based metric based on how much impact you have.
Having great performances on any role increases your performance metric – having higher stats is handicapped on your team with lower quality players on other roles (in which the opposition is on average equal in performance metrics): until you cannot have any impact at all, then things normalizes (since you're virtually no longer having any determinantal impact which needs to be offset for a 'balanced' match)
However, since performing on p1-2 is much more important than other roles to net a win (who cares if your p5 is chain feeding late game if your p1 is stomping – but not reversed) – it is virtually impossible to climb the ladder exclusively on these roles, which also makes p1-2 even more sought after.
Since p1-2 needs to meet minimum criteria for the bracket to net a win, not be far below other carry/ mid players in the bracket
You can steadily climb the ladder far longer if you always have modest p3-5 stats – but it is a reserved playstyle for the 'team encouragement ape' style of player/ Slacks.
This would ultimately matter less if mmr earning/ losses was adjusted for impact metric – and not just a smurf detection threshold.
"besides the forced performance metric there is also game matchmaking manipulation based on how often you play the game. Hence why is better to win some, change account, come back after 3-4 weeks then play again; as the metric most likely gets reset due to inactivity. It's all bullshit as it is truly a matchmaking that has nothing to do with skill but more to addiction to play more and grind."
first game of the day is always mediocre players; at best.
Source: https://www.reddit.com/r/DotA2/comments/s0wrep/team_equal_performance_based_metrics_is_unfair/