100 squats a day, every day, with money on the line.

The squat is the most studied bodyweight movement in the literature, and the bodyweight version is one of the safest loaded patterns a desk-bound adult can practice. Stuart McGill's spine-load work at the University of Waterloo makes the case that an unloaded squat with a neutral spine is a legitimate restorative pattern, not a risky one — provided depth is honest and the lumbar spine stays out of flexion at the bottom. Brad Schoenfeld's 2016 and 2017 meta-analyses on volume and frequency suggest that splitting daily volume into multiple submaximal exposures produces equal or better adaptation than one grinding session, especially for novices and for people whose ceiling is recovery rather than effort. Pavel Tsatsouline argued the same thing without the data in "The Naked Warrior" (Dragon Door, 2003): grease the groove, never go to failure, treat the movement as a skill. For a 100-rep daily target this maps cleanly — 10 sets of 10 between meetings is biomechanically trivial, and the posterior chain and hip-flexor decay that James Levine's NEAT research blames on sedentary work gets quietly reversed in the background. The behavioral failure mode, however, is not biomechanical. It is motivational. The squat needs no equipment, no commute, no warm-up — and that frictionlessness is precisely what makes it skippable. There is always a later. Kahneman and Tversky's prospect theory (Econometrica, 1979) gives the answer: a loss looms about twice as large as an equivalent gain, so attaching a real stake to the daily target inverts the asymmetry. The reps stop competing with Slack and start competing with five dollars walking out of your account toward an animal shelter. Charles Atlas sold this exact insight by mail order in the 1920s under the name Dynamic Tension; the modern version is just better instrumented.