Explanation:
first a quick check that we really deal with 100 events (experiment results).
14+19+16+16+17+18 = 100
perfect.
so, based on this particular die its own probabilty distribution of the 6 different possible outcomes shows that there might be some imbalances inside the die, or its outside dimensions are not exactly the same, or simply the process of rolling the die is not neutral (not fully random).
in any case, it seems there are some deviations from the normal expected behavior of 1/6 per number.
so, the observed experimental probabilty of rolling a 1 is
(as always, desired cases / total cases)
14/100 = 7/50 = 0.14
which is a little bit less then 1/7.