Card game stats - Need to explain to someone how it's not really relevant.

So I play this card game called keyforge.

The game is played by two with random decks of cards with 3 houses each (List of the houses in the graph).
Two new houses have been included in the game : Dinos and Star
The following statistics are the win rate by house of the preview matches (I don't have the actual sample info).

Purely based on these stats, my friend beleives this to be proof that the new houses are "too strong".

To me, there are statistical anomalies that make it a graph that you can't trust as proof of anything: The decks with one or two of the new houses are being played the most (because of novelty and to try them out) which makes the stats appear wrong, because the other houses are not played as much and thus have more chances to be on the "losing" team.

Can anyone help me explain this? Am I wrong?


To recap: The game is played by 2 players
Each deck has any combination of 3 houses , but the players decide which deck they will take (Since two of the houses are new, decks with these houses tend to be played more)
Both players can have the same house combinations.