Does anyone else feel like the percentage of games that are decided by luck before the Turn 4 shuffle or the Turn 7 shuffle is just too high? I feel like it's a little more than 50% of all games where the players are of reasonably equal skill, and this just seems like a flaw.
So far as I can tell, the single largest luck factor in a single game is just: who goes first? Last word on this that I remember suggested a P1 win rate of about .57 averaged over all games. Even if not quite that high now, it's high enough to be remarkable, and I believe all serious matches require equal number of first player starts.
What I find most interesting about Dominion is that, despite the obvious role that chance plays in shuffle luck *and* the relatively small absolute number of "controversial" or "non-book" decisions per game, good players end up with so much better results than poor players. As best as I can tell, a good part of this is because you end up "owning" some of your mistakes for the rest of the game. So, if I buy a card that really doesn't work well with my deck, once per shuffle I will either be playing a card that I wish were something different, or not playing that card and eating the opportunity cost.
I have a hunch that one "metric" that may well do a pretty good job of separating better from weaker players is the number of turns that a card *could* have been bought, but nothing was bought. Or the number of times a more expensive card could have been bought, but a cheaper card was bought instead. Not that this is an automatic strategy, but somebody who does this probably is really thinking about their deck, and if their thinking is any good at all, they are likely to be a pretty decent player.