Truth Table

Answer: CHERRY TOMATO

By Ben Coukos-Wiley

To solve the puzzle, find both solutions to the logic puzzle. The best place to start is the self contained logic loop of Green-2, Green-3, White-1, Scarlett-1, and Mustard-3. But this isn't the only way to go: assigning an arbitrary statement as true or false and following the line of implication will usually lead to either a solution or a contradiction. At the end of each solution you should have determined the truthfulness of all statements and assigned every person a fruit. We then convert each person's four statements into a number by interpereting it as four-digit binary, then we use that number to index their fruit. Reading around the table in the order statements are given in gets us:

Solution 1

Person	Fruit	S1	S2	S3	S4	Index	Letter
Mr. Green	FREESTONE PEACHES	T	T	F	T	13	C
Miss Scarlett	CHAMPAGNE GRAPES	F	F	T	F	2	H
Colonel Mustard	EUROPEAN CANTALOUPE	F	F	F	T	1	E
Professor Plum	ASIAN PEAR	T	F	F	T	9	R
Mrs. Peacock	BLOOD ORANGE	F	T	T	T	7	R
Mrs. White	DANCY TANGERINE	F	T	F	T	5	Y

Solution 2

Person	Fruit	S1	S2	S3	S4	Index	Letter
Mr. Green	EUROPEAN CANTALOUPE	T	T	F	F	12	T
Miss Scarlett	FREESTONE PEACHES	F	T	T	T	7	O
Colonel Mustard	CHAMPAGNE GRAPES	F	T	F	F	4	M
Professor Plum	ASIAN PEAR	F	F	F	T	1	A
Mrs. Peacock	DANCY TANGERINE	F	T	T	F	6	T
Mrs. White	BLOOD ORANGE	F	T	F	F	4	O

Each gives half the answer, and putting them together gets CHERRY TOMATO - a fruit you wouldn't want to find in your bowl of fruit salad.

Author's Notes

The first thing I knew about this puzzle was that I wanted to use fruit salad as a theme. I'm an avid D&D player, and I wanted to reference one of my favorite adages: "Intelligence is knowing a tomato is a fruit. Wisdom is knowing a tomato doesn't belong in a fruit salad." I eventually combined it with an idea I'd been formulating about a logic grid puzzle with liars and truth tellers using the tenuous thematic connection of truth being not all that it seemed.

My original conception of the puzzle was that there would be four solutions, each extracting a statement about the answer to the puzzle. By finding out which were true and false a solver would be able to get enough clues to guess CHERRY TOMATO. The biggest problem was that in order to get useful statements I would have had to include fourteen people and fourteen fruits, making the puzzle ridiculously bloated. Nevertheless, I was determined to see it through.

This version of the puzzle was a "scale model" of that idea to prove it was possible. But after seeing it, Jonah told me it was better to just keep the smaller version, and I decided he was probably right. Even with this smaller version, lots of teams had trouble. We place a lot of that blame on the fact that the flavortext did not sufficiently clue two solutions, but even with that clue solvers found the puzzle much trickier than I anticipated.

I think a big reason that this puzzle wasn't balanced as well as it could have been is that after my original idea wasn't in the cards, I wasn't really as passionate about writing the puzzle. I made a few minor tweaks, but kept the logic exactly as it was in the first version while I lavished puzzles like A Crossover with rewrites to get the details absolutely perfect. Having the quality of a puzzle be so dependent on how invested I am in it is a flaw in how I approach writing, and one I'll have to improve on in the future.

This puzzle also shows that testing doesn't fix everything. None of our testers raised specific red flags about this puzzle, leaving us with the impression it was fine. We should have realized that "fine" isn't the same thing as "good", and worked to make a version of the puzzle that testers were actually enthusiastic about.