There are some choice terms for this kind of bug that could get you banned from this forum. If it is a consistent bug each time you start fresh it would seem to me to be an ordinary problem, as Pete said an initialization problem, or at least a bug in a section of code that is used only the first time through. It should not be hard to look for one-time code and contrary to what you think it is not that rare. Typically you don't test for these types of startup tasks because they tend to be classified as non-functional requirements, but for a thorough test session that is a critical aspect of fitness for use. Rather than to find a name for the type of bug, find the name of the programmer and get it fixed.
Is it repeatable? Can you reload the software and make it do it again? Then yes it is an initialization defect. With out analyzing it further there is no way to say it's a race condition. Usually a race condition occurs erratically and can not be predicted.
If it is not repeatable We would write it up that way and eventually Product Support would be assigned to keep an eye on it in the field as a known non-reproducible defect.
Success is the ability to go from one failure to another with no loss of enthusiasm.
~ Winston Churchill ~