Such statistical funny business would never happen in vehicle safety reporting, but it happens routinely in mutual fund performance reporting. Its called survivorship bias and in the mutual fund industry it happens through the merger of funds - the performance of the poor performing fund is lost and the only high performing fund performance survives in future reporting.
Research studies have shown this practice overestimates true fund performance. So beware of glowing mutual fund performance reports that outwit and outplay fund buyers by hiding the poor performance of those funds voted off the island.
PS - often you see a disclaimer on performance reports that past performance may not be indicative of future performance. In the case of survivorship bias we can take this a step further and say that "past performance may not even be indicate of past performance!"