Potential ways to get past the 66 bottleneck: 1) Try it in the winter. I theorize (without as much data, obviously) that "summer projects" are more likely to be drama filled than something Jan-Mar. 2) Consider 2.5 or 3 classes. You'd still end up with 150+ companies per year. I know it would sort of screw with the current schedule, but there could be some overlap, although at the cost of partner happiness. Maybe have the .5 session be split over a longer period of time, with interviews happening slightly offset from S and demo day slightly offset from W. The longer period would be useful for certain kinds of startups. 3) Invest in a human cloning startup. Once there are 80 clones of each partner, scaling becomes easier. This may take tens or hundreds of years to be effective.