A few things jump out to me. Despite it's lack of any real engine, there are a couple of intriguing paths to take.
The simplest is going to be Masq/BM. It's quite strong, and is very consistent. Even a sub-optimally played deck can get 4 Provinces in 13-15 turns.
The other option is the interesting one. It would consist of a Venture stack with Tunnel, Hoard, and Masquerade. Perhaps Moneylender would be strong as well, but I think that Masq would fill the trashing role well enough. On 3-4$ you would pick up 2 Masquerades, and after that Silver or Tunnel, depending on how long the game has gone. 5$ would be Venture, and 6$ would be a single Hoard, then Gold. Picking up extra green with the Hoard hurts even less than normal due to the Venture, so don't be stingy!
Here you open 5/2, which is really unfortunate. I think the best option is to just bite the bullet and pick up a Masquerade right away. Treasury is extremely weak here, and will never be better than Venture; as there's no way to capitalize on the bottom half. BM isn't helped by it at all, and that's what this board is. On turn 8 you pick up a Gold, which should have been a Hoard. You take both Scheme and Moneylender over Silver, which should be avoided in BM.
The main reason you lost here is that you didn't have any economy. You end the game with just one Silver and one Gold, which even with the Treasuries just isn't enough to buy anything. On these sorts of boards where there isn't an engine you have to prioritize treasures over actions more often than you might expect. Practice makes perfect!
