ATH 2005:

Armchair Treasure Hunt 2005

Title: The Fox Hunt
Setter: Paul Coombs
Themes: Fox Hunting, Past ATHs

Paul Coombs writes...

Some Notes on the Hunt

During the debate about the abolition of foxhunting I occasionally attempted a joke about the Armchair Treasure Hunt being banned, or forced to follow a scented sock across the country. For some reason, this jest always failed to raise even the slightest smile. But I will have my revenge, oh yes - anyone who fails to laugh at my jokes gets punished. Punished! Do you hear! (cue sound of maniacal laughter from the North London area).

Another idea that went into the pot was that of basing a Hunt on the old text-based Dungeons and Dragons computer games where, having defeated some monster, you were rewarded with a choice of directions in which to continue - "there are exits to the north and the west" and so on. If you don't know what I'm on about then ask your dad.

Then, of course, there was the fact that this was the 20th Hunt. From the moment it was decided that I should set this competition, I realised that XX was a clear pointer to having two boxes. Indeed my thinking ran along those lines until very late in the day when I finally decided to have two halves of a puzzle that fitted together. The inspiration for this was that TV advert for Audi in which someone puts together two pieces of microfilm covered in dots to reveal "Vorsprung Durch Technik".

And finally came the idea to set each stage of the Hunt as a tribute to previous Hunts and their setters, while sneakily taking all their best ideas and recycling them as mine. Slowly, it all started to come together...

I have always been disappointed that the ATH has not evolved away from the 'lists of questions surrounded by clipart' format. There has always seemed so much more potential when the only 'givens' are that a box is buried somewhere and there are some clues to its whereabouts. And maybe even those elements are not sacrosanct. So I thought and thought until I finally came up with a brilliant and totally original idea: web pages that consisted of lists of questions surrounded by clipart. But at least a web page opens up some additional possibilities for sounds, movies and links, albeit with the likelihood of some technical cockup disabling the whole thing.

But what about the questions? I polled my quiz-setting friends to see if they could come up with some that did not instantly yield when googled. Although they did unearth a few (that I will doubtless set in some form one day), I then had to ask how the heck anyone could solve such problems. But if there were not to be any questions then there had to be puzzles. And I prefer this anyway. In the early days of the ATH, when we were trying to reproduce Masquerade, the questions were only introduced because we could not emulate Kit Williams' puzzle-filled artwork. I know some people like the questions, but there should be more to the ATH than just spending hours hunched over a search engine - it's far better if those hours are spent in trying to crack a web encoding/password mechanism.

So, once I had decided to have a series of pages each of which revealed the password to unlock the next, there was the problem of how to implement it. Although I suspected that there would be commercial products that could encrypt html, I was determined to do it myself, even though I have not written any actual code since computers were made of valves (far superior to these modern gizmos in my view - you could hear them working and they kept the room warm). Undeterred, I taught myself Javascript and eventually did produce a workable encryption scheme (the hard part is making sure the password itself is concealed). But I was never quite happy with it and eventually turned to a commercial product, just adding a few twists of my own devising.

However, I was still fairly sure that at least some teams would crack this. The very nature of these things means that the decryption algorithm must be visible, so it should be possible to extract the passwords somehow. And doubtless for every commercial product there is something on the web somewhere that will defeat it. So I needed to design the Hunt so that it didn't matter too much if you knew all the passwords - you might have some advantages, but you would still have plenty of work to do to find the treasure. In the event, some teams did successfully 'cheat' in this way (which in fact I don't regard as cheating), but as far as I can see, their methods only worked because the passwords were so short.

As to the location, I was determined from the start to revisit Christmas Common, burial-place in the very first Hunt. An initial reconnaissance revealed something I had forgotten, which is that the name of the pub there is...The Fox and Hounds! The rest was easy, merely taking several hundred hours of effort to determine the puzzles and draw some pictures.

I decided to include a red herring (a good term given the hunting theme) in the form of clues to the village of Cold Christmas in Hertfordshire. This was not just because I expected the passwords to be cracked, but also because it is impossible to encrypt the graphics files (or at least not without a great deal of coding that would have caused no end of problems). And it is really easy to download the entire directory of graphics from the ATH site and have a look at them. So I added some extra pictures in the form of a map of the Cold Christmas area and some spurious directions. Also some encrypted web pages that referenced these, in order to foil the password-crackers. Despite all this, nobody even mentioned the actual village of Cold Christmas in their answers. So a lot of effort wasted and many points unclaimed.

Placing the Hunt on line enabled me to correct a few mistakes as and when they were discovered by various teams. This confused some people, who found that password lengths had changed, or that puzzles were slightly different. I implemented these changes without publicity not (just) out of embarrassment but because it was hard to announce them without giving information away. For example, if I had said that "Rillian" was now "Rilian" (which is the correct spelling by the way) it might have given the whole thing away to a team who had not yet solved that puzzle. However, people who worked from printed copies not only missed these changes but also failed to spot the 'hotspot' clues, for which I apologise, albeit with some reservations.

Of course there is much that I wish I had done differently. The "xenotime" and "xiphias" clues should have been harder, and the final instructions should have been divided into consecutive halves not as part-words. As a result, the treasure was found too early. The passwords should have been longer in order to foil brute-force attacks. And there were too many errors in spelling and transliteration, which are important in these puzzles because everything is taken as a clue (and sometimes is). On the plus side, I did get away from the usual format, and shown that a web-based Hunt is possible. Some more entries, particularly from new Logica teams, would have been nice, but I am encouraged by the interest shown by non-company people in the USA.

If you visited the backup site you might have been intrigued by what else you found on Revelations is a novel written by myself and Lu Zurawski. Free copies can be obtained from Lu, who is a current employee of Logica.

Many thanks to Mark Abbott for helping with an experiment, and for acting as my representative within Logica in order to liaise with the Sports & Social Club. Also to Elaine Bodenitz for help with the publicity, and to Jacqui and Safina for helping on burial day. And finally, a moment of respect for Miles Whitehead, who loved these competitions but who sadly died last year. He is much missed.

I hope you enjoyed this year's Hunt. There will of course be another in December, to be set by Steve Hames. In the meantime, there will soon be a social evening and prize-giving to which I will be inviting everyone who entered. I look forward to seeing you there.

Uncle Paul