Lamb and a kitten are playing a chase game with the following rules:
Steps | Probability |
---|---|
1 | 1/12 |
2 | 1/12 |
3 | 1/6 |
4 | 1/6 |
5 | 1/4 |
6 | 1/4 |
We model this as a Markov process where the state is the current distance between them.
First, compute the expected steps the kitten moves per turn:
E[X] = 1×(1/12) + 2×(1/12) + 3×(1/6) + 4×(1/6) + 5×(1/4) + 6×(1/4)
= 1/12 + 2/12 + 6/12 + 8/12 + 15/12 + 18/12
= 50/12 = 25/6 ≈ 4.1667 steps
Each turn, the expected distance reduction is:
E[reduction] = Lamb's steps + E[X] = 3 + 25/6 = 43/6 ≈ 7.1667 steps
Define E(d) as expected turns needed when distance is d:
E(d) = 1 + Σ P(X=k) × E(max(d - (3 + k), 0))
We compute E(d) recursively from d = 1 to d = 10:
Distance (d) | Calculation | E(d) |
---|---|---|
1-4 | Any move (3+k ≥ 4) covers distance | 1 |
5 | 1 + (1/12)×E(1) | 13/12 ≈ 1.0833 |
6 | 1 + (1/12)×E(2) + (1/12)×E(1) | 7/6 ≈ 1.1667 |
7 | 1 + (1/12)×E(3) + (1/12)×E(2) + (1/6)×E(1) | 4/3 ≈ 1.3333 |
8 | 1 + (1/12)×E(4) + (1/12)×E(3) + (1/6)×E(2) + (1/6)×E(1) | 3/2 = 1.5 |
9 | 1 + (1/12)×E(5) + (1/12)×E(4) + (1/6)×E(3) + (1/6)×E(2) + (1/4)×E(1) | 253/144 ≈ 1.7569 |
10 | 1 + (1/12)×E(6) + (1/12)×E(5) + (1/6)×E(4) + (1/6)×E(3) + (1/4)×E(2) + (1/4)×E(1) | 97/48 ≈ 2.0208 |
E(10) = 1 + (1/12)×(7/6) + (1/12)×(13/12) + (1/6)×1 + (1/6)×1 + (1/4)×1 + (1/4)×1
= 1 + 7/72 + 13/144 + 1/6 + 1/6 + 1/4 + 1/4
= 1 + (14/144 + 13/144 + 24/144 + 24/144 + 36/144 + 36/144)
= 1 + 147/144 = 1 + 49/48 = 97/48
The expected number of turns until the kitten catches Lamb is:
E = 97/48 turns ≈ 2.0208 turns