Colin Mitchell
Let
be the probability
that an event occurs on the first in a sequence of trials. We want to find
, the probability that it occurs on the nth trial, with the probability
that it occurred on
the (n - 1)th
trial, and with the probability
, that it did not occur on the (n
- 1)th trial.
So we can model this with the linear difference equation
.
So if we rewrite this as
,
we can solve it. If we substitute
, we get
.
So a root of this is
, which is multiplicity one.
So we can form the solution as a combination of the general and
particular solutions. Our general
solution is of the form
.
The particular solution is
![]()
Combining these, we get
.
Since we know
, we can find c,

So our final equation becomes
![]()