There is a mathematical theory that is a particular favorite of mine. As roughly stated, the problem is as follows: you are looking to hire a new employee, you have 100 candidates to interview but (and here’s the constraint) you have to hire the winning candidate on the spot at the end of their one interview. Thinking this through, the first candidate may be the best of all the 100, but if you hire the first candidate on the spot you never get to see any of the others (and how do you know that all the others aren’t even better?). Alternatively, if you don’t hire an early candidate that you think is excellent … you might end up kicking yourself because you never get anyone of similar quality and you might be forced to settle for a less good candidate later on (in extremis, you may have to hire candidate number 100 no matter how good they are, because you have no other choice). So what do you do? How do you get the highest probability of hiring someone good?

Is anyone still with me? Excellent. The answer (apparently – I no longer have the maths to prove this myself) is that you interview just under one third of the candidates without hiring anybody, then hire on the spot the next person who is better than all the candidates you have seen so far. And if nobody is as good, hire candidate number 100.

Err, hang on, relevance please? This is a pretty wild digression even for you, James. So…

We are on the outskirts of Tibet, although still officially outside the Tibetan Autonomous Region. The countryside is mountainous and starkly beautiful, monks and monasteries abound and the smells of juniper incense and the squeak of prayer wheels are a constant reminder of where you are. And we are on the lookout for a Buddhist Thangka painting. There are a whole load of art galleries in towns that we are going to visit – once – over the next couple of weeks. We are heading up to Lhasa, which is of course the historical home of all that is excellent in Tibetan art, but history gets pretty convoluted in this part of the world and this may no longer be the case. And we can’t be sure that we won’t end up having to buy Thangka candidate 100 on our last day in Tibet. (see what we have done there? The theory is actually pretty widely applicable – one of the reasons I like it so much)

First stop: Labrang monastery in Xiahe. The monastery itself is perched amidst a small Tibetan enclave at the top of a typically charmless, rapidly growing Chinese town. As a first taste of Tibetan culture it is a dilly – deeply chanting monks, dark mysterious shrines, more deities than you can shake a prayer wheel at and a kora pilgrimage around the monastery walls that doubles as a pleasant stroll for a pair of altitude-acclimatizing westerners.

Next stop: Tongren. And it is here that we have our mathematically resolved dilemma. Tongren (cookie cutter Chinese hole in the ground medium-sized town – don’t bother) has a famous Thangka painting school attached to the Wutun-Si monastery about 6km out of town. And the Thangkas are great. Really great. But are we really about to buy a large Tibetan Buddhist painting and then carry the bloody thing the 1,000+ miles to Lhasa, where there must be more Thankga shops? We ummed and ahhed, we looked at dozens of paintings, we haggled, we negotiated hard … and we bought an awesome painting of a particularly fierce looking bloke on a horse (we discovered later that our fierce looking bloke is in fact a she. And it’s a mule. To be precise, her name is Sri Devi, and she is one of the only protector deities that you offer beer and wine to instead of holy water. Result!)

I would like to say that the mathematics played a part in our decision, and maybe it did. More likely, however, we were so stung by our Tashkent Experience that we would have kicked ourselves hard enough to draw blood had we missed another early opportunity to pick up something high quality and ended up spending another five days hunting for second best. Anyway, we picked up our painting, slid it inside a cannily acquired chunk of sewer pipe and headed off to Lhasa with our fingers crossed that we would never see anything as beautiful again.

Those of you who know me personally and professionally may be surprised to learn that insurance was not my first intellectual interest. That dubious honour goes to engineering – four happy years at Cambridge studying everything from the performance of jet engines to the design of transistors and all that falls in between. And a topic that has sprung to mind recently has been metal fatigue: the process by which a metal body is stressed repeatedly (although not quite to breaking point) until it suddenly gives way without warning well below its design limits. This process leaves a distinctive fracture: bright, shiny and beautiful where the repeated stress has worn away at the edges of the body, then matt, dark and ugly where the final catastrophic failure occurred.

We have just crossed the Torugart Pass. Wildly scenic, this pass winds its way from the rugged south of Kyrgyzstan through the snow-capped Tian Shan mountains to the far western reaches of China. It is remote and deserted, and sees relatively few foreigners due to the extraordinary bureaucratic procedures required to get across it. No fewer than five separate checkpoints, with gun-toting soldiers at each checking your documents, luggage checks, x-rays, form filling and long waits (including one spent standing alone outside the closed gates of China – in case you are wondering, they are high and desolate with five-pointed communist stars and lions on them). You need two drivers, two cars (one Kyrgyz, one Chinese) and all the right documents, and it is an expensive pain in the arse. And you can guess where this post is going, yes? The metal fatigue analogy is all a build up to some spectacular bureaucracy-induced sense of humour failure, right?

Wrong. Lucy and I were actually pretty chilled about all the bureaucracy, as we had employed a specialist travel agent and had managed our expectations extremely hard before we set off. We literally (and by “literally”, I once again mean “literally”) sat back and enjoyed the view. The Big Issue was that once we had got over the pass and were speeding down the road to Kashgar … the back wheel of our car fell off.

Nobody was hurt; nobody was killed. The wheel was – very, very, very fortunately – trapped by the brake disk and the brake pad and merely got stuck at an angle to the car, thereby bringing us to a screeching halt. Indeed, it wasn’t immediately apparent that the back axle had sheared completely off until the driver and I removed the wheel and stuck our heads under the car. There I saw the distinctive (textbook even) pattern of shiny and matt fractures across the inch-thick metal shaft and had a cold sweat moment when I realized what had happened. The driver was still making reassuring noises about “problem with the brakes” when we left him, picked up by the only other tourists who were crossing the pass that day. How the driver got his car picked up we don’t know, and we will never know – you see, we agreed to pay him the $170 for the Chinese side of our trip once we reached Kashgar and, while we made it in one piece in our third car of the day, we have never seen him since.

Some things work in PNG … and some things don’t. We have been having some telecommunications difficulties recently with our rinky-dink global go-anywhere SIM cards, and have also been seriously starved of internet access (well, internet access below $15 a day, and that actually works when you do pay the money). So, what to do?

Well, the answer is: (warning, Geek alert)

• Buy a prepaid local SIM card from Digicel PNG (these are absolutely everywhere, perhaps surprisingly)
• Set it up in my iPhone (so far, so good, and there are instructions for this)
• Subscribe to a prepaid data service
• “Fudge” (from first principles / guesswork / a little white hat inspiration) the Digicel internet tethering APNs to allow the iPhone to act as a data access point
• Physically wire it to my laptop (I don’t have an iPhone 4, so I can’t do this wirelessly)
• Use Connectify on the laptop to set up a wifi hotspot to help out any other tourists in need (and possibly recoup some of my initial SIM card investment…)

And there you go. 100MB of data? 20 Kina (about $10). The feeling of smug satisfaction for having worked out something like this while out in the bush? Priceless!

Caption: Checking the blog comments, while bird of paradise spotting at the Kumul Lodge in the Highlands of Papua New Guinea.

I’m definitely not the first person to use a big fluffy air mattress in a tent. I’m certainly not the first person to fry the electric inflator of the mattress by running too high a voltage through it in an attempt to recharge it extra fast. I’m probably not the first person to jury rig an improvised air mattress inflator using … erm … a car exhaust pipe. However, I am probably one of a select few to do all of the above having woken up that morning in a five star hotel in Las Vegas.

We are in Joshua Tree national park, having driven from the Aria Hotel (cunningly yield-managed to an extraordinarily cheap price on a Sunday night); across the Mojave Desert (which has NO gas stations, for those of you who are thinking of setting out from Las Vegas with a half empty tank); along one of the few remaining portions of the original Route 66; and into the hills of Joshua Tree, all playing obligatory U2 at wildlife-scaring volume. It is meant to be the last hurrah of our flowery tent – due to be thrown away when we reach Los Angeles – and I am lying awake rather worried. The tent is cozy and snug, and we have a double air mattress full of exhaust fumes.

[The Tent’s Last Hurrah (although we eventually relented, and posted it back to the UK)]

Blowing it up actually worked surprisingly well, with Lucy gently gunning the accelerator on request and me clasping the valve over the car exhaust as if a good night’s sleep depended on it. The mattress leaks, however, and exhaust fume asphyxiation in a tight tent isn’t how I intend to go (skydiving head down into an active volcano, for those who are curious). Thinking it through very carefully, however: a slow leak, large tent vents, a fresh breeze and mandatory catalytic converters make it safe, and we sleep tight.

Joshua Tree is famously inhospitable. How the early settlers got their tent pegs in I don’t know. We sleep extremely well under a clear starlit night, then get up at 6:30am and trek for three and a half hours from through the scrub in the scorching heat, down Lost Horse Trail, past an abandoned gold mine and finally up to a view of the San Andreas fault. World class hiking, but we are somewhat national-parked-out at this point having been to Zion, Grand Canyon and Bryce in the last few days, and we are overdue in San Diego for running water, electric light and good sushi.

[Lucy doing her finest Joshua Tree impression – instructions: please wave your hands in the air like you just don’t care!]