Sie sind auf Seite 1von 1

BOMBS OVER LONDON

From June 1944 to March 1945 during World War II, Germany launched a total of 9,251
V-1 flying bombs buzz bombs - against England. Of these, only 2,419 made it to their
intended target areas and, of these, 535 struck South London. Did the bombs strike at random, or
did they have advanced aiming ability?

If they fell at random over South London, then the pattern of hits would follow a
Poisson distribution!

South London was divided into 576 regions of equal area, and the number of hits in each
region determined. Since 535 bombs struck 576 regions, the mean number of hits per region was
535
= = 0.929 .
576

Does the number of hits per region follow the pattern of a Poisson distribution?

Number Probability Observed Relative Expected


of Hits on a of this Number of Frequency of this Number of
Region Number Regions Number Regions
0 0.395 229 0.398 227.5
1 0.367 211 0.366 211.4
2 0.170 93 0.167 97.9
3 0.053 35 0.061 30.5
4 0.012 7 0.012 6.9
5 0.002 1 0.002 1.2
6 or more 0.001 0 0 0.6

Sample Calculations:
e 0.93 (0.93 )2
1. P[2 hits on a region ] = = 0.170
2!
93
2. Relative frequency of 2 hits = = 0.167
576
3. Expected number of regions to be hit exactly twice: since there were 576 regions and
the probability of two hits is 0.170, the expected number of regions to receive two
hits is
np = 576 0.170 = 97.9

Conclusion: Since the observed relative frequencies are each very close to the corresponding
probabilities given by the Poisson distribution, the fit of the Poisson model to the data
is very good. Or, since the numbers of regions receiving each of the various numbers of
hits are each very close to the expected number of regions, the Poisson model fits the
Thus one can conclude randomness of hits rather than an advanced aiming ability.

Das könnte Ihnen auch gefallen