I have 2 types of computers A and B in a data center. Each type has multiple computers. Hardware components in these computers keep failing. I collect data for about a year on how many parts fail.
Total computers of type A -> 80
Total computers of type B -> 65
Total failures in type A -> 50
Total failures in type B -> 36
If there are around 50 total failures in different computers of type A and 36 total failures in different computers of type B, How can I represent them as percentages?
Would it be right to just do 50/80 and 36/65 to get the failure rate? I ask this because out of 50 the failures, a single computer might have had multiple failures. If that is the case, would it be logically correct to just calculate (Number of failures in type A)/(Total number of computers of type A)
If the above method is indeed correct, and suppose my failures in a particular year go up to 100, would it still make sense to report 100/80 as the failure rate for type A?