The accident stats need to be presented along with usage stats to be of any real value.
They are. If you read back in the thread you'll find links to the usage stats, i.e. accidents per 100 million km travelled.
[EDIT] Oops, no it was in the
Cycle to work info thread, but the same PDF I've linked to.
"
Table 9 (page 106) gives casualty rates per 100 million vehicle kilometers.
For 2006 there were 3.1 deaths per 100 million cycling kilometers, 53 killed or seriously injured per 100 million cycling kilometers and 349 casualties (of any severity) per 100 million cycling kilometers.
For cars it was 0.3 deaths per 100 million cycling kilometers, 2.3 killed or seriously injured per 100 million driving kilometers and 29 casualties (of any severity) per 100 million driving kilometers.
To put these in context:-
If I commuted by bike every day (12km each way) I'd do about 5000km a year. It would take me 20,000 years to rack up 100 million kilometers, or 6451 years to reach the distance where there is, on average, one fatality.
For any casualty (down to slightly injured) whilst cycling it's still one per 286533km. That's 57 years worth of my commute.
"
Or, if I commute for another 34 years I've got a 1 in 189 chance of being killed. I'll take those odds any day, especially when compared to the big killer diseases.