She was a cute young girl in a full Halloween costume that had something to do with the game Fortnite.

I don’t know the game, but she was mostly purple, clearly was not dressed as Barney the Dinosaur, and Fortnite was the obvious guess this year for any costume I couldn’t otherwise recognize.

What I did recognize was what she was her discussion of risk and reward. “Just wait,” she pleaded with her gang of friends as they were getting ready to head to the next home, “I am trying to figure out the best return on my investment.”

I don’t know if this sixth-grader truly knows what return on investment is, but she was trying to determine the “right choice” in my annual “Trade or Treat” Halloween exercise. Whether it was her discussion of investing, other kids trying to calculate the risk or just children thinking about the money that is the underlying basis of most ordinary, day-to-day decisions, the talks proved to me that offering kids a choice of cash or candy is worth the cost and makes for a particularly fun time both for the giver and the recipient.

I have been doing something different for kids in or above the third grade for three years now, and this year I had children arrive at my house saying that it was their “favorite in the neighborhood.”

As detailed in this column a month ago, my efforts – initially spawned by a different type of cash-or-candy campaign from the National Financial Educators Council – try to teach financial literacy and decision-making while keeping Halloween the fun and frivolous time I remember it being from my childhood.

What is clear not only from my experience but from those of the many readers who told me they tried their own version of trade-or-treat on Halloween is that candy is dandy, but kids mostly will take the money, honey.

Halloween 2018 was my most complicated exercise, and it produced the most interesting results; I’d like to think the kids learned something beyond “Hey, some guy in the neighborhood is giving away money on Halloween.”

Each year, I have stuck to the basics of what I did prior to offering children cash for candy, namely making three pieces of fun-sized candy – worth roughly 12.5 cents each – the reward for coming to my door. That’s also the only choice that really young kids get.

In 2016, however, kids in third grade and up were told that instead of their three pieces of candy, they could draw a coin envelope. Each of the 50 envelopes I made up that year had at least one quarter in it; the “jackpot” was $5.

Children took all 50 envelopes; four age-eligible kids chose candy. Two kids who would have taken the money were left with candy because I ran out of envelopes.

In 2017, the choices were the three pieces of candy, an envelope with 25 cents and up to $3, or they could let me pick a random piece of fun-sized candy from their bag in exchange for an envelope containing 50 cents minimum up to a $5 grand prize.

All 48 children who took an envelope played for big money. Again, four kids stuck with candy.

This year, age-eligible children again had three options:

1) Three pieces of fun-sized candy;

2) Trade one piece of fun-sized candy picked randomly from their bag for an envelope with at least 50 cents to a maximum of $5. Half of the envelopes contained the 50-cent minimum.

3) Trade three fun-sized candies randomly drawn from their collection to pick from envelopes where any winner would have at least $1, with prizes running up to $10. Half of the envelopes, however, were empty, except for the small paper recapping the choice the children had been given.

Armed with that information, the children effectively were making risk-reward decisions. They could guarantee themselves some money – and go home with more candy – by taking the middle option, or they could risk a few pieces of candy that might otherwise be consumed by their parents or siblings for the chance to play for big money.

By the time the night was over, 43 kids played for big money, with 22 others taking the safer choice and 10 sticking with candy (yes, attendance at Halloween was up this year; add in the little kids and nearly 100 youngsters stopped at my door).

But here’s where the results get particularly interesting.

The total amount of money at stake in the 50 big-money envelopes was $57; in the small-money envelopes, it amounted to $56.50. Thus, if all of the envelopes had been played, kids were choosing roughly the same average result, no matter how much candy they gave me.

Here’s where luck comes in: 23 of the 43 big-money players got empty envelopes. After tallying the money left behind, it turned out that the children took home $43, or an average of $1 per person.

By comparison, the 22 small-money players took home $28, or slightly more than $1.25 per person. Several parents have told me since how happy their kids were making a buck or two having chosen the middle ground.

In an ideal world, the children would discuss the game with their parents – asking the adults what they would do and maybe talking about bigger stakes like lottery tickets and more – learning lessons about risk, reward, choices and the value of money. In a less-than-ideal world, the rumor will spread that I gave out silver dollars, my Halloween crowd will swell [again] and the kids who come to my door expecting money will play without any thought to risk, reward or anything but greed.

My costs for the night were roughly 85 bucks: $71 in cash/coin, a few bucks for coin envelopes and the “money pumpkin” by my door, with less than $10 for candy. It’s probably twice as much as I would have spent just buying candy for nearly 100 visiting costumed kiddies, but it was worth every penny.

Chuck Jaffe is a nationally syndicated financial columnist and the host of "MoneyLife with Chuck Jaffe." Reach him at itschuckjaffe@gmail.com.