There are many factors that are used to make a decision about the platform on which to deploy a system. These include criteria such as risk, ease of maintenance, cost, etc. Those areas that are not researched at the time of choice are usually skipped because those that are making the decision have already set their opinions as to the "facts" in that area.
Unfortunately, many of those assumptions no longer track to what the tracked, objective data now shows.
One area that this now includes is the area of staffing. Several of those beliefs center around assumptions on the cost of mainframe staffing. The typical "knowledge" includes suppositions as to the number of support FTE, the required training of such personnel, availability of needed staff, and the cost of those positions.
This paper examines the data backing any position on mainframe staffing and balances those assumptions against the cold, hard facts of today.