Pablo's Armchair Treasure Hunt

Armchair Treasure Hunt 2008

Title: The Royal Parks
Setter: Tim North
Themes: Royal Parks, London Landmarks

Tim North writes...

Some Notes on the Hunt

I volunteered to set an ATH about two and a half years ago because, at one of the post-ATH celebrations, Pablo had remarked that setters were thin on the ground. I figured that there is an ATH in everyone and so set out to write one. I had competed in eight at that time and reckoned that was enough experience to be able to set one. It is, however, harder than it seems.

Quite early on I had decided upon my objectives for my ATH. I wanted it to be approachable for beginners but taxing for the experienced ATHers. ATH is notoriously difficult to comprehend for newcomers. That meant ensuring that there were plenty of questions that could be answered fairly easily, hence drawing competitors in to the quiz. I also wanted avoid clues and constructs that would automatically give the advantage to experience rather than skill.

The use of apparently simple questions was my mechanism for making ATH approachable, but at the same time I had to make it more difficult than it seemed. That was why getting the correct answer was important - for example in question 24 you had to get Skipper rather than the obvious Barbie. Having started to use that mechanism I realised that I couldn't do too many like that [other than the difficulty of finding suitable questions] because the N/S/E/W would become too obscure. I hope I did enough to make it statistically obvious.

I decided on the theme quite by accident. I had been in London on the London Eye and looking over London decided that in that vast city, full of its own little treasures, there must be a suitable place to hide a tiny little box. I bought a fold-out map and started to look at all the green areas.

Some of the 'squares' and small parks looked like they could be good hiding places, but then knowing what London is like at night I was doubtful that it could hold a treasure box secret for a month. The parks looked like they could be large enough to hide a box - but a visit to a couple of them showed that they are quite well looked after and that burying a box could be difficult and if found could lead to a potential security alert. The park regulations are also quite strict about permitted activities and my brother's offer to visit me in jail put me off that idea - at least for a while.

I had wanted the theme to be 'London', however I decided that although I personally find London interesting and a worthy theme other hunters may not share my interest and that many may feel disadvantaged by not living or working in London. If, in a few years, someone wants to set a hunt themed on London then please do - there's plenty of material!

Eventually I decided that the Royal Parks could provide good subject material and a theme and that the precise location of the treasure didn't have to be inside the park and could be just outside the jurisdiction of the park regulations. Technically, where I hid the box is also covered by the same regulations but I didn't discover that until it was too late.

The treasure was buried just outside of Bushy Park, in a small park. My selection of the eight parks was initially limited by those open 24x7, like Bushy and Green. I didn't want treasure hunters queuing outside locked gates waiting for them to open. Hiding the treasure outside removed that restriction but by that time I'd found the small park and a suitable fallen tree.

Where was the 'L'? There wasn't one because defacing anything in a Royal Park was out of the question and would definitely have landed me in trouble.

In order to find the treasure you needed to reassemble the 49 5x5 letter grids back in to a large Wordsearch grid. This was my favourite part of the hunt. It took me a considerable amount of time to assemble. These 5x5 grids had their letters randomly rotated to make it more difficult, and the filler letters were selected to provide an even spread. The words contained within the Wordsearch were London landmarks and therefore finding those would help you look for fragments of words. The words were distributed over the grid to confirm the unique solution. Some of you may have noticed that the letters N, S, E and W appeared in the top, bottom, right and left middle squares respectively at the edge to further confirm orientation.

I wrote Java code to handle the generation of the random filler letters, the random rotation, and the assembling of the final grid using image handling routines. It also followed the route and provided me with confirmation that it was valid and circular. The use of Java was a good investment as it enabled me to quickly regenerate the grid, which I had to do many times especially in the last few days when I discovered a couple of careless spelling mistakes in my answers.

I didn't know how long it would take hunters to reassemble the grid so tried it out on myself. After 1.5 hours I'd done most of it, albeit with a couple of mistakes, so reckoned that it was achievable. I was close to putting the landmarks reading in all four directions, and even diagonally, but felt that might be a step too far. I think that the difficulty of that element was about right in the end, which is pleasing and I enjoyed seeing the photographs of pieces of paper carefully laid out on coffee tables. One of the first ATHs that I did was one that involved cutting it in to pieces and reassembling it, and was, I felt, a good way to get involvement from the rest of the family who can understand what is required for that element of the hunt without needing to have spent hours looking at it.

Following on from my original theme of 'London' I considered whether I could overlay a grid over a map of central London and use the answers to skip from square to square, or travel along a roads and using the answers a guide to which direction to turn at junctions. Sadly, London doesn't have quite the same road structure as places such as New York and is really just a sprawling mass of roads. It did however spark my idea of having answers that started with an N/S/E/W.

I then spent the next twelve or so months collecting suitable questions, jotting them down in my little notebook, and ripping out articles from The Metro. Much of this material was later discarded because it was impossible to set a suitable clue, or the answers were too ambiguous, or in the case of those collected in 2007 irrelevant to a quiz at the end of 2008.

It took me several weeks to get out of the habit of noticing words that start N/S/E/W!

I think that if there one was mistake that I made in setting ATH it was choosing such a narrow subject. Having eight parks restricted me to either eight core pages, unless I took the big jump to two pages per park, and each park only holds a finite number of interesting statues and landmarks.

I also learnt that introducing randomness into ATH is not necessarily a wise thing to do. In ATH everything on the page has significance and deliberately breaking from that to add the randomness of letter rotations, the pseudorandom cyphertext and two of the keys did, I think cause some competitors to waste time trying to figure out the meanings. Looking back on it I should have used the letter rotation angles to encode something that signified the end of that trail, e.g. Happy Christmas!

As I've said in the solution, I tried to stress the importance of Christmas Day. I suppose that I was also trying to stress the unimportance of Christmas day as well. ATH is what Christmas is all about.

It was not my intention to put in any 'red herrings' as these are conventionally signalled in some way that experienced hunters will spot and the virgins will fail to - hence widening the gulf between the two which I had wanted to narrow.

The only red herring that I'll admit to is that the first three answers spell out WWW, which was a deliberate attempt to get the experienced hunters thinking back to ATH2003 and hopefully sending them down a short, blind, alley. One cut of the questions/answers spelled out a plausible website address, though once I realised it led to a possibly dubious website I resequenced the questions.

I don't know whether it is common among setters, but setting ATH led me to a degree of caution [paranoia]. Could my ciphers be cracked in seconds using online tools? Was anyone hacking in to my laptop? Would the use of PDF reveal my filenames? Would the use of Visio reveal the rotation angle of the letters? Would the treasure be found on the first day? Would I be arrested for wandering around the undergrowth areas of the parks?

I buried the treasure on Sunday 14th December, taking along one of my children as a plausible cover story for my odd activities. I had a minor worry because since locating the site some plastics barriers had been erected nearby and it looked like park maintenance were about to perform some long overdue repairs, such as removing that untidy fallen tree. Checking back on the site a week or so later there was no sign of any works in progress and I figured that close to Christmas would be an unusual time for them to start a new project.

The use of a pseudorandom key sequence gave me a good defence against a last-minute change of site, which I had already used to my advantage. That was that the final instructions could be changed in a matter of minutes, feeding it through the Java encryption code, split in to segments and pasted in to the pages.

On 15th December I published the quiz. Some of you may have noticed that the file was time stamped less than half an hour before the noon launch. I printed out the final copy and removed a picture from the final page that showed the area around the burial site. It just looked too obvious. Perhaps it wouldn't be to someone looking at it for the first time but I'd been looking at these pages for months and it was now saying 'It's here!!' From the submissions I see that quite a few of you visited the now infamous bus stop D and would quite possibly have speculatively started looking (in fact I think that one person did but didn't spot it).

I regret that the published copy was lacking in image quality but my concerns that the PDF could be decoded using the professional Adobe tools (that I didn't have access to and therefore couldn't check) meant that I had to effectively publish the final copy as a set of images.

There were two teams who I am most grateful to for allowing me access to their Wiki/forums. As has been said before, the setter publishes the quiz and then it all goes quiet. By having access to some feedback I was able to assess whether I had made any mistakes or anything that needed clarification. I'd made a couple of spelling mistakes, but the main error was "On page 4, TSAE should read TSEW". Strangely, I've never been very good at remembering my North, South, East, West, and had put the wrong direction. Stupid.

There was however an unfortunate side-effect of having access to the Wiki/forums. That was that I felt that I had to monitor them a lot more than I had planned, so my quiet relaxing Christmas was spoilt by the addictive need to see what the teams were up to.

The first Saturday was the day that I had expected the treasure to be found, if not before. I was correct and received a call from Chris Andrews, whose team had located the treasure at about 8am. I then expected phone calls from the other teams who I felt would be shortly behind - but as it turned out none of the other teams were able to find the treasure using the clues.

With a day or two to go leading up to the 19th January and only one finder I was considering options. Should I extend the deadline, which could disadvantage Chris's team in the Best Solution category and may also unfairly upset some of the other team's schedules? I did, however, want the second and third positions to be filled. After much consideration I concluded that accepting the solutions as they stood at the original deadline was the fairest thing to do and extended the deadline purely for awarding second and third to treasure. I know that led to some frustration at the time, however looking back on it I still consider that to have been the best thing to do.

Marking. I had not envisaged quite how close the entries would be. I had put a few items in (for example the 'War Games' map) that I had hoped would be differentiators but more teams solved more clues than I had expected. I therefore had to be very strict on how I marked the solutions and what constituted a mark. I dropped some of the questions from the scoring because I was unable to draw a definite enough line between correct/incorrect.

I hope that you all enjoyed the hunt this year. I'm sorry that marking it took longer that I had planned. I needed a rest from ATH, having been living for ATH for almost two years (and especially in October-January), of a few weeks after I accepted the solutions and getting back in to it was difficult especially as work picked up considerably following Christmas.

Would I set an ATH again? Possibly.