The sock begins with a square to form the ball of the foot. Two quarters of that square are put on a waste-yarn holder while the other two quarters are decreased to form closure at the top of the toe. The decreases are worked in four sections to end with just 2 stitches in each section. These remaining 8 stitches are gathered with yarn threaded on a tapestry needle.
At this point, the sock shape isn’t readily apparent.
The two sections of held stitches are then returned to the needles and worked in a bias tube with decreases worked along the top of the foot and increases worked along the bottom of the foot.
When knitting socks, I typically work with four double-pointed needles–three to hold the stitches and one to knit with. Hovever, the “square” format of this sock construction makes it easier for me to hold the stitches on four needles (one needle per section) and knit with a fifth.
So far, the sock has an odd shape. The cast-on “point” at the ball of the foot looks suspiciously like a heel. I think blocking is going to be a key step in this pattern.