Tim Lambert (deltoidblog AT gmail.com) is a computer scientist at the University of New South Wales.
I was able to find some statistics on homicides in South Australia (in "South Australian Historical Statistics" Vamplew, Richards,Jaensch and Hancock)
Unfortunately, they only cover the period 1921-1979, but we can use them to see if gun control, introduced in SA in 1929, had any effect.
We can use the adjacent state of NSW as a "control". The SA homicide figures include "manslaughter by driving" and the NSW figures do not, so they are not strictly comparable, but before 1940 they were only two or three cases of manslaughter by driving each year in SA.
The second last column is the probability (two-tailed Student t test) that random variation could cause the difference. The last column is the Mann-Whitney U statistic (significant at 5% level if it outside the range 20-70).
Ave 1921-1929 Ave 1930-1939 Prob chance? U SA 3.4 1.5 0.00005 88 NSW 1.5 1.5 0.7 44.5
Conclusion: homicide rates in South Australia fell significantly following gun control in 1929, while NSW rates did not change.
(BTW, while my source has attendance figures for every cricket game in Adelaide since 1892, it doesn't have figures on population broken down by age.)
The data:
Homicide rates per 100,000 population
NSW SA
1921 1.8 3.4
1922 1.6 3.0
1923 1 2.0
1924 0.9 5.1
1925 1.1 3.7
1926 1.6 4.0
1927 1.6 3.0
1928 1.9 2.8
1929 1.7 3.9
1930 1.7 1.2
1931 1.6 2.4
1932 1.3 1.6
1933 1.5 1.1
1934 1.4 1.8
1935 1.8 0.9
1936 1.4 1.9
1937 2 2.5
1938 1.2 1.5
1939 1.3 0.6
1940 1.7 0.8
