The point being that in the 30 years since Chernobyl we have had no serious accidents involving any modern nuclear reactor models and one serious accident involving a reactor older than Chernobyl. We may just have too little data, but at least so far the data all points to the argument that modern reactors are safer being reasonably likely to be true.
The only reason we don't have clear evidence either way is that the risks even with the older models is very low.
The incidence of accidents with any reactors is very low.
The risks associated with those accidents are extremely high.
I'm just relaying the argument as I understand it.