Explanation:
so, we had
4+3+2+2+6 = 17 experiments
I assume the spinner has 5 equality sized fields with different colors.
so, normally we would assume that in average each color has the same probability : 1/5 (desired results over totally possible results).
the experiments with the actual spinner give us now observational data and probabilities.
after all, a spinner (as a die, or a roulette wheel, or ...) can have slightly different probabilities for the individual outcomes based on design and manufacturing.
so, based on these results we see the actual probabilities like this
red 4/17
blue 3/17
green 2/17
yellow 2/17
purple 6/17
so, the observed probability shows us a probability of blue in the next spin to be 3/17.
please note, that after that next (18th) spin all the observed probabilities have to be updated depending on the actual result.
so, if the result of spin 18 is indeed blue, the probabilities for spin 19 would change to
red 4/18
blue 4/18
green 2/18
yellow 2/18
purple 6/18