Understanding a probability inequality from Billingsley's Converegence of Probability Measures

51 Views Asked by At

In the below proof, I cannot figure out how we get the final inequality. Namely, how do we show that $$P[\tau_i^n - \tau_{i-1}^n \ge \delta' | \tau_q^n < m] \ge 1-\eta/P[\tau_q^n < m]$$ from (16.29)?