Of course, being an engineer I must point out Yet Another Problem (TM) with testing all the M&M's (TM) from a bag against the previous winner using your fingers. The strong M&M (TM) heats up due to body temperature, and loses much of its structural strength, as the material inside the shell softens. Secondly, the thumb provides a much larger surface area to the device under test (DUT) than the forefinger provides to its DUT. This will lead to skewing in the results, though it might be offset by the additional heat provided to the thumb DUT because of not only a larger surface area, but a greater blood flow, and more heat transfer to the DUT. Use of a vice, pliers, or other room temperature object is therefore suggested.

Obviously, though, each M&M (TM) has surface and internal imperfections which will affect the results. You may ask, "Well DUH! That's what we're determining with this test!" Let me propose, however, a situation wherein this is detrimental to the test. If a strong M&M (TM) has a minute dimple just the right shape of the other M&M (TM) then it would provide more surface area to the second, weaker M&M. By spreading the stress over a larger surface the result would be the weaker M&M (TM) sapping power from the strong M&M.

To combat this each M&M (TM) must face similar surfaces on both sides. This can be accomplished with another metal blank between the two M&M's.

The evolutionary aspects of such dimples, however, are beyond the scope of this discussion.