Normal mode is different from ranked mode (according to the game developer's explanation) only in that it does not explicitly display player rank, but still uses similar ranking and matchmaking systems. My studies find that players tend to treat normal games as casual game play while apply a competitive mindset to ranked games; and to many players who mostly play ranked, playing normal games is to take a break.
But casual play in normal mode is not entirely judgment-free. On the one hand, it is associated with a strong culture on what characters players should play. The seeming consensus is that one should not select characters which they have already played frequently in ranked (and are supposedly good at), because if they "try hard" in norms, they have "no life." In my recent normal games I've observed quite a few incidents where players were mocked for choosing their mains in norms, and especially so if they lost. On the other hand, players still want to win, and blame teammates who don't do well. So the ideal scenario of normal games is that a player picks a character they have never played before, but still do well in game and finish game with bragging rights. Not saying this doesn't happen in ranked. It does! I've had numerous games where players bragging about playing a character the first time in a ranked game.
There could be many reasons why a player can perform very well with an unfamiliar character. They might have an overall, deep understanding of the game and in-game mechanics. They might be lucky against a weak opponent. They might have a strong team... But this makes the game feel, random. Should game design encourage or discourage this phenomenon? How to balance learning curve and power of a particular character, to cater to both average game players and pro gamers?