Spire Shadows (non-market variant: draw in 2 out of 3 games till round 3)

In my prior Spire Shadows market fetch version I was dedicating 14 market fetch cards to reliably play Spire Shadows .
This attempt instead places four spires in the deck and adds a few draw accelerators.

Drawing a spire on the initial hand of 7 has a 33% probability - not bad for a start. A redraw with 6 cards adds a 19% chance, which could be considered if the opening first hand of 7 cards doesn't have any other draw accelerators. The raw chance to draw a spire in the first 10 cards drawn is 44% (7 cards opening hand + 3 cards drawn until 3rd turn).

I keep four of each Cull the Deck and Scheme to search for a spire in the top of the deck. There is a 75% chance to draw at least one of them amongst the 10 cards drawn until turn 3. So for the 56% chance (100% - 44%) that I didn't draw a spire till round 3, I would at least have a 75% chance to peek into the next 3 cards for a spire at 18% chance. So the compounded likelihood to find a spire by round 3 is 44% + (100% - 44%) * 75% * 18% = 44% + 8% = 56%. The true chance is higher because I might find multiple draw cards that can be chained, going 2nd gives me an extra card draw, and Scheme would allow me to peek into top 4 cards, and I optionally could redraw my initial hand in the first place. It ballpark boils down finding a spire in 2 out 3 games early enough to matter.

The units in this deck are almost highlander unique. A common theme (given their low attack value) is that they are defensive which helps to stall the game. There is plenty of room to optimize focusing on top picks.
Some of the units and spells would also need to be sacrificed for removals which I haven't included yet.