The stochastic process is a simple random walk if
where each step in the walk corresponds to the outcome of a Bernoulli trial, so is independent of all other steps.
What is the probability that a simple random walk, started at will reach the value 3 before it reaches the value 0?
So by Corollary 1.3.3:
Now we calculate the probability of the gambler’s ruin.
Consider two simple random walks, and , with defined as in 2.2.1, and defined by
Assume . Define the events
Finally let and . Then
The and processes differ only in that the probability of moving up is equal to the probability of moving down, and vice-versa.
Let the rv be the time at which first becomes 0, i.e. the time to ruin, starting from = 1. We note the possibility that may not have a proper distribution, i.e. one which sums to 1. In fact that is what we wish to determine, the value of
If this is less than 1 we say that has an improper distribution.
Firstly note that for we need to be odd (as we need one more downward move than upward move). Writing we get
where is the number of paths for the stochastic process that start with , end with and with for . The probability of each of these paths is as each of these paths involves upward moves, and downward moves.
[To see this we can look at the example , and draw all realisations of the simple random walk that end in :
There are 5 such paths, and each individual path has probabilty . So we get ]
Now define to be the time at which first becomes 0. By a similar argument we have
the only difference being through the interchange of and due to the different probabilities of up/down moves. Finally we note
∎
Let be a simple random walk (RW) starting from . Define the event:
i.e. or or …. Then
Note that and are same as and but still we prefer to use separate notations. This is one aspect of the Gambler’s ruin problem. The Gambler starts with £1 and plays a succession of games, winning or losing £1 at each game with respective probabilities and . We say that ruin occurs when the gambler has no £s left, i.e. when = 0 for the first time. Play must then stop.
Our proof consists of two parts, and uses Lemma 2.2.3. In part (a) we calculate , the probability of ruin if , in terms of . In part (b) we use part (a), and Corollory 1.3.3, to obtain a quadratic equation for . Lemma 2.2.3 is then used to show which root of this quadratic equation we should take.
Part (a). Consider a more general walk starting from ; let
and let . We call this ‘Ruin starting with £k’. We now show that .
Consider a sequence of games. The gambler starts with £k in her pocket, takes out £1 and plays until that is lost. She then takes out another pound and does the same. And so on, playing games until she has no £ left i.e. is ruined. This is the only way in which ruin can occur. Each game is independent of the others and is lost with probability , so the required probability is . A graphical representation for is
Formally, define the event to be the ‘passage from the state to reach the level for the first time’, and call this time the rv . Then let be the ‘passage from to reach the level for the first time’, and call this time . And so on until for the first time which is the time of ruin . Then
because all these events are independent. The reason for their independence is that the outcome of the second game is not affected in any way by the loss of the first £, and so on. The events , , … correspond exactly to the event of ruin for the game starting with just £1, so have the same probability .
The lengths of time of each game, which we call , , …, are therefore also independent, a fact we use for a later result.
Part (b). Condition on the outcome of the first step in the RW starting from .
Now use Corollory 1.3.3:
But from part (a), . Also because ruin has just occurred. So
This give the quadratic equation for :
When , . Next assume . Consider a plot of these two roots as function of :
If then because the value is not possible, the value of must be 1. If , then it is not possible to have , otherwise, by Lemma 2.2.3 . So we now know that for . ∎
The expected time to ruin starting from , in the case is
This is only meaningful when is a proper rv with . From Theorem 2.2.4, this holds when i.e. .
Part (a). As before, let be the time at which ruin occurs starting from . Then
where is the length of the first game and so on. Now , , …all have the same distribution as . They are also independent, although that is not neccessary for this result, which is
Part (b). Condition on the outcome of the first step, with the same figure as before:
and use . Now
so that
where is the further time to ruin from . But has exactly the same distribution as and from part (a), , giving
Then so .
It is possible similarly to evaluate , hence and higher moments.
We have already calculated the probability of ruin without using the expression for . Next we calculate the probability of ruin at the -th stage by deriving an expression for .