## 3984 – Odd One Out by Oyler -Setter’s Blog

Posted by Listen With Others on 22 Jun 2008

**A Setter's Blog by Oyler**

I’d just set two puzzles for The Magpie – Quadwrangle and Quadwrangle II which were based on an earlier 2002 Listener puzzle 3968 Twinset by Zag and a Tough Crosswords puzzle in December 2001 Light At The End Of The Tunnel by Googly. In those puzzles there were two grids and two sets of clues and solvers had to decide which clue went with which grid. In QW and QWII I doubled these up and had four 4×4 grids and four 5×5 grids respectively that were paired up in some way. The next logical step was four 6×6 grids. However I decided against this on the grounds that the puzzle would be too big ( at least 20 clues ) to fit comfortably on an A5 page without significant reduction in print size thus annoying those who like myself are visually challenged.

QW and QWII were set in a 3 week period and as I was on a bit of a roll I wanted to continue and wondered about merging Twinset with another Googly puzzle Red Herrings that appeared in the March 2003 issue 6 of The Magpie in which solvers had one grid but two letter/number assignment clues and so had to reject one of them. I decided to move it up a notch and have two grids and three clues one of which would remain unused.

I barred off the two 5×5 grids differently. In the previous puzzles by Zag and Googly the grids were identical however I prefer them to be different as useful information can be obtained from product and quotient clues that can eliminate an entry from a particular grid and so makes it a bit easier. I was about to start the clueing process when it struck me that it was a shame that one clue would be unused and wondered if that could form the basis for a third unseen grid and it would be this grid that solvers had to submit for their entry making it easier for John Green to check.

Obviously each grid had to have the same number of clues and as luck would have it the two grids that I’d barred off just happened to have the same number of bars and unchecked cells as well as the bars in different places. I decided that the third grid would have that also. There were a number of possibilities for the third grid given that no two bars across all three grids were in the same place and I chose the one that had the entry lengths increasing to a maximum then decreasing again as it was easy to explain if that information had to be given. I had hoped to be able to miss that out but that proved not to be possible as I couldn’t solve the puzzle without it!!

Clues were going to be of two types as in the previous puzzles namely number definition clues such as prime, square etc and entry referenced algebraic clues such as 2A, b / D etc. I had used the double duty number definition clues idea in QW and QWII and as this was going to be a Listener puzzle decided to use it again as it makes things a bit more difficult for the solver. For example knowing a fully checked entry is 89 and it refers to a clue Fibonacci, Prime, A x B solvers have to carry the clue over into the other grids as 89 is both Fibonacci and prime. Of course it all resolves itself in the end!

I started with the A across clues first but as the puzzle was set some 3 years ago that’s about as much as I can remember. Sorry!

When setting puzzles that use number definition type clues I make use of a few double sided sheets that contain lists of squares, primes etc that have been photocopied from various sources namely Wells’ Penguin Dictionary of Curious and Interesting Numbers and Jenkins’ Number File ( Tarquin Press ) then cut and pasted in order to get as much information as possible on a sheet then rephotocopied and laminated. When setting Euler’s Spoilers I made up a table for numbers 0 – 99 that detailed what type of number each was and I still use this in a slightly amended form.

Sq = Square. Cu = Cube. Tn = Triangular. Pr = Prime. Fib = Fibonacci. Luc = Lucas. Sm = Smith. Ha = Happy. Lky = Lucky. Ab = Abundant. MP = Multiplicative Persistence – multiply the digits together to get a new number and continue until you get a single digit, the MP is the number of steps required to get to the single digit.

The puzzle took 2 to 3 weeks to set during which ELP, The Nice, Deep Purple, Trace, Ekseption and Triumvirat had a good airing on my hi-fi. For some reason I set puzzles best listening to 60s and 70s prog rock. It must be something to do with the keyboard playing and classical overtones!

## Leave a Reply