Let's talk winrate: correlationa and causation.

These days you'll look at lolking, op.gg, or champion.gg for winrates, and think "oh that champion is op" or "damn that champ is weak." You waddle over to a forum and, I guarantee you, at least a few people will comment saying that winrates mean "nothing". I am here to tell you why that is a wrong stance to take in statistics.

Firstly, every statistic indicates some sort of correlation. When doctors tell you that smokers are some percentage more likely to get lung cancer, it means that across a large population, there is a statistically significant increase in people who get lung cancer in smokers vs non smokers.

The laws of statistics don't change when it comes to League. When one champion wins more often than not across many games, there is a good indication that that champion is somehow strong, that is, that champion influences games positively in its player's favor to a certain extent. It might be because of itemization, masteries, runes, strength of other roles, or even the skill of its playerbase.

A 47% winrate doesn't necessarily indicate whether the champion is strong, but if the only statistic you have is that winrate, there's a good chance that champion is not doing well more often than not. If that same champion has 50% of its playerbase playing it for the first time and losing, then odds are there's a reason so many people want to play a champion despite not being good at it (namely the champion being strong). But if 90% of its playerbase has over 100 games on that champion, and even then they're doing badly, that is a VERY good indication that that champion is weak.

Now if a champion has a 55% winrate, and 95% of its playerbase mains that champion, that champion can't necessarily be deemed overpowered.

Winrate is a good indicator. It is not an end-all statistic, but nonetheless, it isn't useles.