# Red herrings in detective puzzles

OK, so I've written several detective puzzles.

Part 1 turned out to be what I'd consider balanced for difficulty: it had a clearly-correct answer, and took about half a day to be solved.

Part 2 was an unmitigated disaster: it took 20 minutes to solve, and someone else guessed the end-game I had planned.

Part 3 was in between the two - it took about 4 hours to solve.

It's obviously very easy to just carpet-bomb the puzzle with red herrings all over the place, littering it with irrelevant clues. I feel that's a bit blunt though.

The converse is just as bad: if the puzzle only contains relevant information then someone can start anywhere and pretty quickly figure out the answer.

What I'm wondering is given an only-relevant-information detective puzzle, how much distraction should I add? As a supplementary question, how can I judge it to try to strike a balance between the solution and the red herrings?

• In my humble opinion, a great puzzle is difficult to guess not because of red herrings but because it is illusive, and yet once you know the answer, you immediately know it is correct. – Neil Dec 5 '14 at 14:35
• I have the exact same problem with steganographic puzzles. – A E Dec 5 '14 at 16:06
• In my humble opinion, red herrings are the best part of a puzzle, and you should put as many of them as you like, and cackle maniacally when someone barks up the wrong tree. – Kingrames Oct 4 '16 at 1:05

I believe detective puzzles are a bit special, as in that it is in their very nature that there are only tiny amounts of relevant information in what is otherwise a haystack of information. It is all about searching and finding.

However, I think a good detective puzzle differentiates false information in

• just not relevant

• real red herrings. (apparently relevant but misleading)

For the second category, I think a fair detective puzzle should allow the red-herrings to be disregarded by logic. i.e. there should be something in the available information which with some thinking allows a clear conclusion: "This can not be it". Note the difference between the conclusion "No, it can not be because..." and No this is not sufficient proof for..." I think it should be possible, to really rule any red-herring out completely.

As for the question of "how much..." , I don't think there is clear answer to that. I think it's experience and feedback from puzzle-solvers which can (iteratively) help you there to find the right balance.

My own attempt was far too easy...

• The point about making it logically feasible to discount a line of thinking is interesting. Do you mean that it should be possible to know for sure that a line of reasoning can be ignored, or would something else just clueing that and hinting that it's a distraction be acceptable, in your opinion? – Joe Dec 5 '14 at 14:46
• @Joe: Yes, that's what I meant. A line of reasoning which is a red herring should have a logical reason to be known as false for sure. (In a fair and fun puzzle.) – BmyGuest Dec 5 '14 at 15:08