Here is an idea of admittedly limited utility: apply the principles of natural selection to a candy dish.
First, one must acquire a variety of candies of various types. In this particular example, I have ordered them (left to right) from most-to-least desirable. (The rightmost item is a toothbrush, which is the universally-despited dentists’ halloween treat.)
Fig 1: Four types of “candies” in the candy bowl. Although individual preferences may vary, in this case we assume that the overall preferences are as follows: Yellow > Orange > Green > Blue. We will refer to these as “candies” even though the toothbrush is not, strictly speaking, a candy in the traditional sense.
Fig 2: The initial candy bowl is (approximately) equally populated by the four candy types.
Fig 3: The candy bowl after co-workers / children / passers-by have visited it for some amount of time. Note that the yellow candies are heavily depleted, but the undesirable blue ones are still almost all present. (This is because they have a higher resistance to predators.)
After each step, we repopulate the candy bowl with one new candy for each (say) four of a given type. So if there are a total of 12 yellow candies in the bowl, we add three (= 12 / 4) more yellow ones. (This may require purchasing a substantial quantity of new candies.)
Fig 4: After another round of candy acquisition, the yellow and orange candies have almost been hunted to extinction.
Fig 5: The good candies have all been eaten, so now the candies’ natural predators refocus their attentions on the green ones.
Fig 6: Even the green candies are almost extinct now.
Fig 7: The candy bowl endgame consists of a monoculture of blue toothbrushes that no one wants. Success! The blue toothbrush has emerged as the most resistant candy with the highest fitness in this specific environment.
PROS: Could be an interesting experiment to illustrate population changes due to selection pressure.
CONS: Requires purchasing a lot of candy!