Solution
= Solution
With no <censoring> or new entry within the interval, all $r$ individuals can be classified at its end: $d$ have experienced the event and $r-d$ remain event-free. Thus the empirical conditional <survival probability> is
$$
\boxed{\widehat{\mathbb P}(T>t_R\mid T>t_L)=\frac{r-d}{r}=1-\frac dr.}
$$
This assumes $r>0$ and that the observed individuals represent the group whose conditional <survival probability> is being estimated.