The thing is unless we have a systematic survey of people who have stretched a) steel frames b) alloy frames, sub-categorised as to how much the stretch was, the age, use and condition of the bike. Then this is enhanced with similarly segmented data about failures attributed to the stretch ( and...