Here's a clipping from Cornell University's turfgrass writeups, after a three year test:
I personally play things conservatively on cool-season grasses in this area, and don't include Potassium applications as a recommendation for annual plans I write for applications after October 1st. While I'm not sure the evidence is overwhelming, there are plenty of other opportunities in the April - September period to get the soil levels right, and I don't like rolling the dice. I don't think a few percent of Potassium in fertilizers in the Fall is a major issue, but full-blown quantities I avoid, and I recommend avoidance to others.Gray and pink snow mold infestation was assessed twice during spring 2008 by estimating the percent area per plot infected. As observed in previous seasons, the incidence of snow mold was increased at increasing potassium application rates. In fact there was a significant effect of the late season potassium rate on snow mold incidence, i.e., as late season rate increased snow mold incidence increased. There was also an obvious reduction in recovery associated with high seasonal and late season potassium applications.