Not really bro, the only people rerolling and were in the arena doors early on are either people who predicted very well and did it towards the end of 9.0(the people who rushed 2100 on alts before the patch), or people without jobs who can reroll and be ready within a short ammount of time. Unless we can see how many people have overly excess ammounts of time to play, representation is a totaly skewed approach to looking at the meta. Not every holy pally got the pleasure of rerolling to one of the other 4 healers, and if you didnât already gear a MW before 9.1, it reduces the chances of someone playing MW even if they are A tier or better. Not saying MW is, but if it was, you wouldnât see 90% of holy pally players swapping to MW, more like 20%, or 10%, or however many no-lifers can accomplish an early reroll.
Iâve done that very consistently when it comes to discussing the factors that cause a spec to be represented poorly on a leaderboard, you just couldnât understand something tongue in cheek pointing out how silly it is when people draw their conclusions on class strength off of data alone because I know people that post here will look at a website like this and say that their spec really isnât that good.
That was the point.
youâre not going to see many non necro rshamans at the top, or night fae enh. There is a meta in place because it feels the most comfortable.
Also people will switch their class because then people wont play with them. Like imagine how hard it would be to que high rated arenas as a MW monk when they can pick up a rdruid or restro shaman or holy priest.
Look at the checkpvp of high rated players. they all have alts and are heavily invested in the game. So they will level an alt and play a class that is stronger to get a competitive edge.
Itâs much more general than that. You identified one potential confounding variable. The problem is that people think that the data on these sites is due to a causal relationship (spec power => representation), but thereâs no valid statistical way to infer such a relationship from the presented data.
pretty sure those grades are based on the 3v3 bracket. If so, then they would be correct. 
âRepresentation is bad because I donât want my broken OP class to be nerfed. I enjoy having a built in advantage over othersâ. There ya go I fixed it for you.
they see their politicians use stats incorrectly so they figure why not them too
explain then the seasons where WW was strong but was lowly represented?
I miss arguing with you about stats, but my goodness thatâs funny and true.