To estimate the importance of social isolation for individuals of different ages to slowing the growth of the COVID-19 pandemic, we measured the speed of COVID-19 case growth as a function of the proclivity of social contacts, both in general and with older people. COVID-19 case growth speed increased with number of social contacts with people >65 years old, but decreased with average number of contacts across all ages. We introduce the Social Antagonist Hypothesis, suggesting that contacts of younger people with other younger people keep young people away from older people, offering relative protection to the older population more vulnerable to COVID-19 from contagion by younger people. To test this, we correlate the number of contacts of younger people to those of older people, and find a significant negative correlation, as predicted by the Social Antagonist Hypothesis. We then find this negative correlation is due to a negative correlation between young-to-young contact counts and young-to-old ones. We conclude that “Senior Hours” for older populations only may be counterproductive toward the goal of protecting the most vulnerable population.